inst/testdata/standards_imd_anova.R
In pmartR/pmartR: Panomics Marketplace - Quality Control and Statistical Analysis for Panomics Data
# Load functions for calculating IMD-ANOVA standards ---------------------------
source(system.file("testdata",
"imd_anova_standard_fns.R",
package = "pmartR"))
# Load data and prepare omicsData objects --------------------------------------
load(system.file('testdata',
'little_pdata.RData',
package = 'pmartR'))
pdata <- as.pepData(e_data = edata,
f_data = fdata,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
# Create additional f_data objects with different main effects and covariates.
fdata2 <- fdata
set.seed(22250)
fdata2$Level <- c(sample(c("high", "low"), 9, replace = TRUE),
"none", "none", "none")
# Create a vector that has three different levels for Condition.
fdataZ <- fdata
set.seed(720)
fdataZ$Condition <- c(sample(c("zombie", "mutant"), 9, replace = TRUE),
"human", "human", "human")
fdataZ$Gender <- sample(c("F", "M"), 12, replace = TRUE)
fdataZ$Age <- round(runif(12, min = 19, max = 89), 2)
# Copy edata so the names of the samples can be changed.
edata2_2 <- edata
# Change some of the Infection samples to Mock samples.
names(edata2_2) <- c("Mass_Tag_ID",
paste0("Infection", 1:6),
paste0("Mock", 1:6))
# Create additional f_data objects with different main effects and covariates.
fdata2_2 <- fdata
# Update the sample names in f_data.
fdata2_2$SampleID <- c(paste0("Infection", 1:6),
paste0("Mock", 1:6))
# Update the first main effect to account for changing some infection samples to
# mock samples.
fdata2_2$Condition <- c(rep("Infection", 6),
rep("Mock", 6))
fdata2_2$Level <- c("high", "low", "high", "low", "high", "low", "high",
"high", "low", "low", "low", "high")
fdata2_2$Gender <- fdataZ$Gender
fdata2_2$Age <- fdataZ$Age
# Create pepData objects, log the data, add group_DF attributes ----------------
# NOTE: There are three numbers that follow each object's name. The first
# number represents the number of main effects. The second number represents
# the number of covariates. The third represents the number of groups (across
# all main effects).
# One main effect, zero covariates, two groups
pdata_1_0_2 <- pdata
pdata_1_0_2 <- edata_transform(pdata_1_0_2, "log")
pdata_1_0_2 <- group_designation(omicsData = pdata_1_0_2,
main_effects = "Condition")
groupDF_1_0_2 <- attr(pdata_1_0_2, "group_DF")
# One main effect, one covariate, three groups
pdata_1_1_3 <- as.pepData(e_data = edata,
f_data = fdataZ,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
pdata_1_1_3 <- edata_transform(pdata_1_1_3, "log")
pdata_1_1_3 <- group_designation(omicsData = pdata_1_1_3,
main_effects = "Condition",
covariates = "Gender")
groupDF_1_1_3 <- attr(pdata_1_1_3, "group_DF")
# One main effect, two covariates, three groups
pdata_1_2_3 <- as.pepData(e_data = edata,
f_data = fdataZ,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
pdata_1_2_3 <- edata_transform(pdata_1_2_3, "log")
pdata_1_2_3 <- group_designation(omicsData = pdata_1_2_3,
main_effects = "Condition",
covariates = c("Gender", "Age"))
groupDF_1_2_3 <- attr(pdata_1_2_3, "group_DF")
# Two main effects, zero covariates, three groups
pdata_2_0_3 <- as.pepData(e_data = edata,
f_data = fdata2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_0_3 <- edata_transform(pdata_2_0_3, "log")
pdata_2_0_3 <- group_designation(omicsData = pdata_2_0_3,
main_effects = c("Condition", "Level"))
groupDF_2_0_3 <- attr(pdata_2_0_3, "group_DF")
# Two main effects, one covariate, four groups
pdata_2_1_4 <- as.pepData(e_data = edata2_2,
f_data = fdata2_2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_1_4 <- edata_transform(pdata_2_1_4, "log")
pdata_2_1_4 <- group_designation(omicsData = pdata_2_1_4,
main_effects = c("Condition", "Level"),
covariates = "Gender")
groupDF_2_1_4 <- attr(pdata_2_1_4, "group_DF")
# Two main effects, two covariates, four groups
pdata_2_2_4 <- as.pepData(e_data = edata2_2,
f_data = fdata2_2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_2_4 <- edata_transform(pdata_2_2_4, "log")
pdata_2_2_4 <- group_designation(omicsData = pdata_2_2_4,
main_effects = c("Condition", "Level"),
covariates = c("Gender", "Age"))
groupDF_2_2_4 <- attr(pdata_2_2_4, "group_DF")
# Filter IMD-ANOVAly -----------------------------------------------------------
# One main effect, zero covariates, two groups
filta_1_0_2 <- imdanova_filter(pdata_1_0_2)
afilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# One main effect, one covariate, three groups
filta_1_1_3 <- imdanova_filter(pdata_1_1_3)
afilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# One main effect, two covariates, three groups
filta_1_2_3 <- imdanova_filter(pdata_1_2_3)
afilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, zero covariates, three groups
filta_2_0_3 <- imdanova_filter(pdata_2_0_3)
afilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, one covariate, four groups
filta_2_1_4 <- imdanova_filter(pdata_2_1_4)
afilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, two covariates, four groups
filta_2_2_4 <- imdanova_filter(pdata_2_2_4)
afilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Save filter objects ---------------------------------------------------------
# save(
#
# afilta_1_0_2,
# gfilta_1_0_2,
# cfilta_1_0_2,
#
# afilta_1_1_3,
# gfilta_1_1_3,
# cfilta_1_1_3,
#
# afilta_1_2_3,
# gfilta_1_2_3,
# cfilta_1_2_3,
#
# afilta_2_0_3,
# gfilta_2_0_3,
# cfilta_2_0_3,
#
# afilta_2_1_4,
# gfilta_2_1_4,
# cfilta_2_1_4,
#
# afilta_2_2_4,
# gfilta_2_2_4,
# cfilta_2_2_4,
#
# file = "/Users/mart077/pmartR/inst/testdata/standards_filter.RData"
# )
# Assemble ANOVA standards -----------------------------------------------------
Xmatrix_1_1_3 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1,
1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0), dim = c(12L, 4L), dimnames = list(
c("1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11",
"12"), c("(Intercept)", "Groupzombie", "Grouphuman", "GenderM"
)), assign = c(0L, 1L, 1L, 2L), contrasts = list(Group = "contr.treatment",
Gender = "contr.treatment"))
Xmatrix_1_2_3 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1,
1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57), dim = c(12L,
5L))
Xmatrix_2_1_4 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0), dim = c(12L, 4L))
# with interaction effect
Xmatrix_2_1_4_full <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,
0), dim = c(12L, 5L))
Xmatrix_2_2_4 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57), dim = c(12L,
5L))
# with interaction effect
Xmatrix_2_2_4_full <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57, 0, 0, 0, 0,
0, 0, 0, 0, 1, 1, 1, 0), dim = c(12L, 6L))
# Selected Full/Reduced Model
which_X_2_1_4_a <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0))
which_X_2_2_4_a <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0))
which_X_2_1_4_g <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0))
which_X_2_2_4_g <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0))
# main effects: 1; covariates: 0; groups: 2 ---------------
mean_a_1_0_2 <- data.frame(
Mean_Infection = rowMeans(afilta_1_0_2$e_data[, 2:10],
na.rm = TRUE),
Mean_Mock = rowMeans(afilta_1_0_2$e_data[, 11:13],
na.rm = TRUE)
)
pval_a_1_0_2 <- data.frame(
Infection_vs_Mock = a(data = afilta_1_0_2$e_data[, -1],
groups = attr(afilta_1_0_2, "group_DF")$Group)
)
flag_a_1_0_2 <- data.frame(
Infection_vs_Mock = aflag(
grp1 = mean_a_1_0_2$Mean_Infection,
grp2 = mean_a_1_0_2$Mean_Mock,
pvals = pval_a_1_0_2[, 1],
cutoff = 0.05
)
)
astan_1_0_2 <- data.frame(
Mass_Tag_ID = afilta_1_0_2$e_data$Mass_Tag_ID,
Count_Infection = unname(rowSums(!is.na(afilta_1_0_2$e_data[, 2:10]))),
Count_Mock = unname(rowSums(!is.na(afilta_1_0_2$e_data[, 11:13]))),
mean_a_1_0_2,
Fold_change_Infection_vs_Mock = (
mean_a_1_0_2[, 1] - mean_a_1_0_2[, 2]
),
P_value_A_Infection_vs_Mock = pval_a_1_0_2$Infection_vs_Mock,
Flag_A_Infection_vs_Mock = flag_a_1_0_2$Infection_vs_Mock,
row.names = NULL
)
class(astan_1_0_2) <- c("statRes", "data.frame")
attr(astan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(astan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(astan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_total = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_anova = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_anova = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_gtest = c(0),
Down_gtest = c(0),
row.names = NULL
)
attr(astan_1_0_2, "statistical_test") <- "anova"
attr(astan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(astan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(astan_1_0_2, "pval_thresh") <- 0.05
attr(astan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_0_2$e_data[, -1])) /
prod(dim(afilta_1_0_2$e_data[, -1]))),
num_samps = dim(afilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(astan_1_0_2, "which_X") <- rep(0, nrow(astan_1_0_2))
attr(astan_1_0_2, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_0_2$e_data$Mass_Tag_ID,
flag_a_1_0_2
)
attr(astan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
data_1_1_3 = afilta_1_1_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_1_3), Xmatrix = data.matrix(Xmatrix_1_1_3))
group_sampnames <- groupDF_1_1_3[,get_fdata_cname(afilta_1_1_3)]
groupData <- groupDF_1_1_3[group_sampnames %in% colnames(afilta_1_1_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_1_1_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", covariate_names = "Gender")
mean_a_1_1_3 <- get_lsmeans(data = data_1_1_3, xmatrix = Xmatrix_1_1_3, pred_grid = pred_grid, Betas = Betas)
beta_to_mu = pred_grid
beta_to_mu[,4] = 0
beta_to_mu = unique(beta_to_mu)
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
cmat = cmat %*% beta_to_mu
test_values <- get_test_values(afilta_1_1_3$e_data[, -1], Xmatrix_1_1_3, cmat)
group_counts_1_1_3 <- data.frame(
nona_mutant = rowSums(!is.na(data_1_1_3[, c(1, 3:4, 9)])),
nona_zombie = rowSums(!is.na(data_1_1_3[, c(2, 5:8)])),
nona_human = rowSums(!is.na(data_1_1_3[, 10:12]))
) %>%
dplyr::rowwise() %>%
dplyr::mutate(n_grp = sum(dplyr::c_across(nona_mutant:nona_human) == 0)) %>%
dplyr::ungroup()
nona_grps_1_1_3 <- rowSums(group_counts_1_1_3[, 1:3] != 0)
nona_counts_1_1_3 <- rowSums(!is.na(data_1_1_3))
# set mean_a_1_1_3 to NA for groups with zero observations
mean_a_1_1_3[group_counts_1_1_3[,-4] == 0] <- NA
diffs_1_1_3 <- mean_a_1_1_3 %>%
dplyr::mutate(
diff_m_z = Mean_mutant - Mean_zombie,
diff_m_h = Mean_mutant - Mean_human,
diff_z_h = Mean_zombie - Mean_human
) %>%
dplyr::select(diff_m_z, diff_m_h, diff_z_h)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3")
test_stat_1_1_3 <- diffs_1_1_3 %>%
cbind(group_counts_1_1_3) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_m_z = (diff_m_z / C1),
stat_m_h = (diff_m_h / C2),
stat_z_h = (diff_z_h / C3)
) %>%
dplyr::select(stat_m_z, stat_m_h, stat_z_h) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_1_1_3 <- test_stat_1_1_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = pt(
q = abs(stat_m_z),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_mutant_vs_human = pt(
q = abs(stat_m_h),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_zombie_vs_human = pt(
q = abs(stat_z_h),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`row.names<-`(NULL)
flag_a_1_1_3 <- data.frame(
mutant_vs_zombie = aflag(
grp1 = mean_a_1_1_3$Mean_mutant,
grp2 = mean_a_1_1_3$Mean_zombie,
pvals = pval_a_1_1_3[, 1],
cutoff = 0.05
),
mutant_vs_human = aflag(
grp1 = mean_a_1_1_3$Mean_mutant,
grp2 = mean_a_1_1_3$Mean_human,
pvals = pval_a_1_1_3[, 2],
cutoff = 0.05
),
zombie_vs_human = aflag(
grp1 = mean_a_1_1_3$Mean_zombie,
grp2 = mean_a_1_1_3$Mean_human,
pvals = pval_a_1_1_3[, 3],
cutoff = 0.05
)
)
astan_1_1_3 <- data.frame(
Mass_Tag_ID = afilta_1_1_3$e_data$Mass_Tag_ID,
Count_mutant = group_counts_1_1_3$nona_mutant,
Count_zombie = group_counts_1_1_3$nona_zombie,
Count_human = group_counts_1_1_3$nona_human,
mean_a_1_1_3,
Fold_change_mutant_vs_zombie = (
mean_a_1_1_3[, 1] - mean_a_1_1_3[, 2]
),
Fold_change_mutant_vs_human = (
mean_a_1_1_3[, 1] - mean_a_1_1_3[, 3]
),
Fold_change_zombie_vs_human = (
mean_a_1_1_3[, 2] - mean_a_1_1_3[, 3]
),
pval_a_1_1_3,
Flag_A_mutant_vs_zombie = flag_a_1_1_3[, 1],
Flag_A_mutant_vs_human = flag_a_1_1_3[, 2],
Flag_A_zombie_vs_human = flag_a_1_1_3[, 3],
row.names = NULL
)
class(astan_1_1_3) <- c("statRes", "data.frame")
attr(astan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(astan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(astan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_total = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_1_1_3, "statistical_test") <- "anova"
attr(astan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(astan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(astan_1_1_3, "pval_thresh") <- 0.05
attr(astan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_1_3$e_data[, -1])) /
prod(dim(afilta_1_1_3$e_data[, -1]))),
num_samps = dim(afilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(astan_1_1_3, "which_X") <- rep(0, nrow(astan_1_1_3))
attr(astan_1_1_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_1_3$e_data$Mass_Tag_ID,
flag_a_1_1_3
)
attr(astan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_1_3, "data_class") <- "pepData"
tukey_stat_1_1_3 <- test_stat_1_1_3 * sqrt(2)
n_tests_1_1_3 <- rowSums(!is.na(tukey_stat_1_1_3))
suppressWarnings(
tukey_pval_1_1_3 <- tukey_stat_1_1_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = ptukey(
abs(stat_m_z),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
),
P_value_A_mutant_vs_human = ptukey(
abs(stat_m_h),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
),
P_value_A_zombie_vs_human = ptukey(
abs(stat_z_h),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_1_1_3 <- which(rowSums(!is.na(tukey_stat_1_1_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_1_1_3 <- which(!is.na(pval_a_1_1_3[just_one_1_1_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_1_1_3[just_one_1_1_3, ][not_na_1_1_3] <- data.matrix(
pval_a_1_1_3[just_one_1_1_3, ]
)[not_na_1_1_3]
# Convert back to a data frame.
tukey_pval_1_1_3 <- data.frame(tukey_pval_1_1_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
dunnett_1_1_3 <- pval_a_1_1_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_1_1_3 - no_nas) %>%
cbind(test_stat_1_1_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_m_z = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_z),
no_nas),
rep(abs(stat_m_z),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_m_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_h),
no_nas),
rep(abs(stat_m_h),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_z_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_z_h),
no_nas),
rep(abs(stat_z_h),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_m_z, pval_m_h, pval_z_h)
# main effects: 1; covariates: 2; groups: 3 ---------------
data_1_2_3 = afilta_1_2_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_2_3), Xmatrix = data.matrix(Xmatrix_1_2_3))
group_sampnames <- groupDF_1_2_3[,get_fdata_cname(afilta_1_2_3)]
groupData <- groupDF_1_2_3[group_sampnames %in% colnames(afilta_1_2_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_1_2_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", c("Gender", "Age"))
mean_a_1_2_3 <- get_lsmeans(data = data_1_2_3, xmatrix = Xmatrix_1_2_3, pred_grid = pred_grid, Betas = Betas, continuous_covar_inds = 5)
beta_to_mu = pred_grid
beta_to_mu[,4] = 0
beta_to_mu = unique(beta_to_mu)
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
cmat = cmat %*% beta_to_mu
group_counts_1_2_3 <- data.frame(
nona_mutant = rowSums(!is.na(data_1_2_3[, c(1, 3:4, 9)])),
nona_zombie = rowSums(!is.na(data_1_2_3[, c(2, 5:8)])),
nona_human = rowSums(!is.na(data_1_2_3[, 10:12]))
) %>%
dplyr::rowwise() %>%
dplyr::mutate(n_grp = sum(dplyr::c_across(nona_mutant:nona_human) == 0)) %>%
dplyr::ungroup()
nona_grps_1_2_3 <- rowSums(group_counts_1_2_3[, 1:3] != 0)
nona_counts_1_2_3 <- rowSums(!is.na(data_1_2_3))
mean_a_1_2_3[group_counts_1_2_3[,-4] == 0] <- NA
diffs_1_2_3 <- mean_a_1_2_3 %>%
dplyr::mutate(
diff_m_z = Mean_mutant - Mean_zombie,
diff_m_h = Mean_mutant - Mean_human,
diff_z_h = Mean_zombie - Mean_human
) %>%
dplyr::select(diff_m_z, diff_m_h, diff_z_h)
test_values <- get_test_values(afilta_1_2_3$e_data[, -1], Xmatrix_1_2_3, cmat)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3")
suppressWarnings(
test_stat_1_2_3 <- diffs_1_2_3 %>%
cbind(diff_denoms) %>%
cbind(group_counts_1_2_3) %>%
dplyr::mutate(
stat_m_z = (diff_m_z / C1),
stat_m_h = (diff_m_h / C2),
stat_z_h = (diff_z_h / C3)
) %>%
dplyr::select(stat_m_z, stat_m_h, stat_z_h) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
)
pval_a_1_2_3 <- test_stat_1_2_3 %>%
dplyr::mutate(
lg = group_counts_1_2_3$n_grp,
P_value_A_mutant_vs_zombie = pt(
q = abs(stat_m_z),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_mutant_vs_human = pt(
q = abs(stat_m_h),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_zombie_vs_human = pt(
q = abs(stat_z_h),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`row.names<-`(NULL)
flag_a_1_2_3 <- data.frame(
mutant_vs_zombie = aflag(
grp1 = mean_a_1_2_3$Mean_mutant,
grp2 = mean_a_1_2_3$Mean_zombie,
pvals = pval_a_1_2_3[, 1],
cutoff = 0.05
),
mutant_vs_human = aflag(
grp1 = mean_a_1_2_3$Mean_mutant,
grp2 = mean_a_1_2_3$Mean_human,
pvals = pval_a_1_2_3[, 2],
cutoff = 0.05
),
zombie_vs_human = aflag(
grp1 = mean_a_1_2_3$Mean_zombie,
grp2 = mean_a_1_2_3$Mean_human,
pvals = pval_a_1_2_3[, 3],
cutoff = 0.05
)
)
astan_1_2_3 <- data.frame(
Mass_Tag_ID = afilta_1_2_3$e_data$Mass_Tag_ID,
Count_mutant = group_counts_1_2_3$nona_mutant,
Count_zombie = group_counts_1_2_3$nona_zombie,
Count_human = group_counts_1_2_3$nona_human,
mean_a_1_2_3,
Fold_change_mutant_vs_zombie = (
mean_a_1_2_3[, 1] - mean_a_1_2_3[, 2]
),
Fold_change_mutant_vs_human = (
mean_a_1_2_3[, 1] - mean_a_1_2_3[, 3]
),
Fold_change_zombie_vs_human = (
mean_a_1_2_3[, 2] - mean_a_1_2_3[, 3]
),
pval_a_1_2_3,
Flag_A_mutant_vs_zombie = flag_a_1_2_3[, 1],
Flag_A_mutant_vs_human = flag_a_1_2_3[, 2],
Flag_A_zombie_vs_human = flag_a_1_2_3[, 3],
row.names = NULL
)
class(astan_1_2_3) <- c("statRes", "data.frame")
attr(astan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(astan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(astan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_total = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_1_2_3, "statistical_test") <- "anova"
attr(astan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(astan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(astan_1_2_3, "pval_thresh") <- 0.05
attr(astan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_2_3$e_data[, -1])) /
prod(dim(afilta_1_2_3$e_data[, -1]))),
num_samps = dim(afilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(astan_1_2_3, "which_X") <- rep(0, nrow(astan_1_2_3))
attr(astan_1_2_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_2_3$e_data$Mass_Tag_ID,
flag_a_1_2_3
)
attr(astan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_2_3, "data_class") <- "pepData"
tukey_stat_1_2_3 <- test_stat_1_2_3 * sqrt(2)
n_tests_1_2_3 <- rowSums(!is.na(tukey_stat_1_2_3))
suppressWarnings(
tukey_pval_1_2_3 <- tukey_stat_1_2_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = ptukey(
abs(stat_m_z),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
),
P_value_A_mutant_vs_human = ptukey(
abs(stat_m_h),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
),
P_value_A_zombie_vs_human = ptukey(
abs(stat_z_h),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_1_2_3 <- which(rowSums(!is.na(tukey_stat_1_2_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_1_2_3 <- which(!is.na(pval_a_1_2_3[just_one_1_2_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_1_2_3[just_one_1_2_3, ][not_na_1_2_3] <- data.matrix(
pval_a_1_2_3[just_one_1_2_3, ]
)[not_na_1_2_3]
# Convert back to a data frame.
tukey_pval_1_2_3 <- data.frame(tukey_pval_1_2_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
suppressWarnings(
dunnett_1_2_3 <- pval_a_1_2_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_1_2_3 - no_nas) %>%
cbind(test_stat_1_2_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_m_z = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_z),
no_nas),
rep(abs(stat_m_z),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_m_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_h),
no_nas),
rep(abs(stat_m_h),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_z_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_z_h),
no_nas),
rep(abs(stat_z_h),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_m_z, pval_m_h, pval_z_h)
)
# main effects: 2; covariates: 0; groups: 3 ---------------
# no covariates? should just be the observed means.
mean_a_2_0_3 <- data.frame(
Mean_Infection_high = rowMeans(afilta_2_0_3$e_data[, c(2:4, 7:8)],
na.rm = TRUE),
Mean_Infection_low = rowMeans(afilta_2_0_3$e_data[, c(5:6, 9:10)],
na.rm = TRUE),
Mean_Mock_none = rowMeans(afilta_2_0_3$e_data[, 11:13],
na.rm = TRUE)
)
group_counts_2_0_3 <- data.frame(
nona_iHigh = rowSums(!is.na(afilta_2_0_3$e_data[, c(2:4, 7:8)])),
nona_iLow = rowSums(!is.na(afilta_2_0_3$e_data[, c(5:6, 9:10)])),
nona_mNone = rowSums(!is.na(afilta_2_0_3$e_data[, 11:13]))
)
nona_grps_2_0_3 <- rowSums(group_counts_2_0_3 != 0)
sigma_2_0_3 <- dplyr::rowwise(afilta_2_0_3$e_data[, -1]) %>%
dplyr::mutate(
stdev = summary(lm(
dplyr::c_across(Infection1:Mock3) ~ groupDF_2_0_3$Group
))$sigma
) %>%
dplyr::pull(stdev)
nona_counts_2_0_3 <- rowSums(!is.na(afilta_2_0_3$e_data[, -1]))
diffs_2_0_3 <- mean_a_2_0_3 %>%
dplyr::mutate(
diff_high_low = Mean_Infection_high - Mean_Infection_low,
diff_high_none = Mean_Infection_high - Mean_Mock_none,
diff_low_none = Mean_Infection_low - Mean_Mock_none
) %>%
dplyr::select(diff_high_low, diff_high_none, diff_low_none)
test_stat_2_0_3 <- diffs_2_0_3 %>%
dplyr::mutate(
stat_high_low = (diff_high_low /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 1] +
1/group_counts_2_0_3[, 2]))),
stat_high_none = (diff_high_none /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 1] +
1/group_counts_2_0_3[, 3]))),
stat_low_none = (diff_low_none /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 2] +
1/group_counts_2_0_3[, 3])))
) %>%
dplyr::select(stat_high_low, stat_high_none, stat_low_none)
pval_a_2_0_3 <- test_stat_2_0_3 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_high_low),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_none = pt(
q = abs(stat_high_none),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_none = pt(
q = abs(stat_low_none),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_none,
P_value_A_Infection_low_vs_Mock_none)
flag_a_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_high,
grp2 = mean_a_2_0_3$Mean_Infection_low,
pvals = pval_a_2_0_3[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_none = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_high,
grp2 = mean_a_2_0_3$Mean_Mock_none,
pvals = pval_a_2_0_3[, 2],
cutoff = 0.05
),
Infection_low_vs_Mock_none = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_low,
grp2 = mean_a_2_0_3$Mean_Mock_none,
pvals = pval_a_2_0_3[, 3],
cutoff = 0.05
)
)
astan_2_0_3 <- data.frame(
Mass_Tag_ID = afilta_2_0_3$e_data$Mass_Tag_ID,
Count_Infection_high = unname(
rowSums(!is.na(afilta_2_0_3$e_data[, c(2:4, 7:8)]))
),
Count_Infection_low = unname(
rowSums(!is.na(afilta_2_0_3$e_data[, c(5:6, 9:10)]))
),
Count_Mock_none = unname(rowSums(!is.na(afilta_2_0_3$e_data[, 11:13]))),
mean_a_2_0_3,
Fold_change_Infection_high_vs_Infection_low = (
mean_a_2_0_3[, 1] - mean_a_2_0_3[, 2]
),
Fold_change_Infection_high_vs_Mock_none = (
mean_a_2_0_3[, 1] - mean_a_2_0_3[, 3]
),
Fold_change_Infection_low_vs_Mock_none = (
mean_a_2_0_3[, 2] - mean_a_2_0_3[, 3]
),
pval_a_2_0_3,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_0_3[, 1],
Flag_A_Infection_high_vs_Mock_none = flag_a_2_0_3[, 2],
Flag_A_Infection_low_vs_Mock_none = flag_a_2_0_3[, 3],
row.names = NULL
)
class(astan_2_0_3) <- c("statRes", "data.frame")
attr(astan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(astan_2_0_3, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none")
attr(astan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_total = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_2_0_3, "statistical_test") <- "anova"
attr(astan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(astan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(astan_2_0_3, "pval_thresh") <- 0.05
attr(astan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_0_3$e_data[, -1])) /
prod(dim(afilta_2_0_3$e_data[, -1]))),
num_samps = dim(afilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(astan_2_0_3, "which_X") <- rep(0, nrow(astan_2_0_3))
attr(astan_2_0_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_0_3$e_data$Mass_Tag_ID,
flag_a_2_0_3
)
attr(astan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_0_3, "data_class") <- "pepData"
tukey_stat_2_0_3 <- test_stat_2_0_3 * sqrt(2)
n_tests_2_0_3 <- rowSums(!is.na(tukey_stat_2_0_3))
suppressWarnings(
tukey_pval_2_0_3 <- tukey_stat_2_0_3 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_high_low),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_none = ptukey(
abs(stat_high_none),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_none = ptukey(
abs(stat_low_none),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_none,
P_value_A_Infection_low_vs_Mock_none) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_0_3 <- which(rowSums(!is.na(tukey_stat_2_0_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_0_3 <- which(!is.na(pval_a_2_0_3[just_one_2_0_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_0_3[just_one_2_0_3, ][not_na_2_0_3] <- data.matrix(
pval_a_2_0_3[just_one_2_0_3, ]
)[not_na_2_0_3]
# Convert back to a data frame.
tukey_pval_2_0_3 <- data.frame(tukey_pval_2_0_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(5)
dunnett_2_0_3 <- pval_a_2_0_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_0_3 - no_nas) %>%
cbind(test_stat_2_0_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_high_low = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_high_low),
no_nas),
rep(abs(stat_high_low),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_high_none = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_high_none),
no_nas),
rep(abs(stat_high_none),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_low_none = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_low_none),
no_nas),
rep(abs(stat_low_none),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_high_low, pval_high_none, pval_low_none)
# main effects: 2; covariates: 1; groups: 4 ---------------
data_2_1_4 = data.matrix(afilta_2_1_4$e_data[, -1])
covariate_names = colnames(attr(attr(afilta_2_1_4, "group_DF"), "covariates"))[-1]
main_effect_names = attr(attr(afilta_2_1_4, "group_DF"), "main_effects")
group_sampnames <- groupDF_2_1_4[,get_fdata_cname(afilta_2_1_4)]
groupData <- groupDF_2_1_4[group_sampnames %in% colnames(afilta_2_1_4$e_data),]
groupData[,main_effect_names] <- lapply(groupData[main_effect_names], function(x) factor(x, levels=unique(x)))
groupData <- groupData %>%
dplyr::left_join(afilta_2_1_4$f_data)
# no continuous covariates
pred_grid_red_2_1_4 = get_pred_grid(groupData, c("Condition", "Level"), "Gender")
pred_grid_full_2_1_4 = get_pred_grid(groupData, c("Condition", "Level"), "Gender", as.formula("~Condition*Level+Gender"))
cobra <- run_twofactor_cpp(
data = data.matrix(data_2_1_4),
gpData = groupData[,c("Group", main_effect_names, covariate_names)],
Xfull=Xmatrix_2_1_4_full, Xred = Xmatrix_2_1_4,
pred_grid_full = pred_grid_full_2_1_4, pred_grid_red = pred_grid_red_2_1_4,
continuous_covar_inds = numeric(0)
)
group_counts_2_1_4 <- data.frame(
nona_Infection_high = rowSums(!is.na(data_2_1_4[, c(1, 3, 5)])),
nona_Infection_low = rowSums(!is.na(data_2_1_4[, c(2, 4, 6)])),
nona_Mock_high = rowSums(!is.na(data_2_1_4[, c(7, 8, 12)])),
nona_Mock_low = rowSums(!is.na(data_2_1_4[, c(9, 10, 11)]))
) %>%
dplyr::ungroup()
nona_grps_2_1_4 <- unname(rowSums(group_counts_2_1_4 != 0))
nona_counts_2_1_4 <- unname(rowSums(!is.na(data_2_1_4)))
mean_a_2_1_4 <- data.frame(cobra$lsmeans)
colnames(mean_a_2_1_4) <- paste0("Mean_", unique(groupData$Group))
diffs_2_1_4 <- mean_a_2_1_4 %>%
dplyr::mutate(
diff_ih_il = Mean_Infection_high - Mean_Infection_low,
diff_ih_mh = Mean_Infection_high - Mean_Mock_high,
diff_ih_ml = Mean_Infection_high - Mean_Mock_low,
diff_il_mh = Mean_Infection_low - Mean_Mock_high,
diff_il_ml = Mean_Infection_low - Mean_Mock_low,
diff_mh_ml = Mean_Mock_high - Mean_Mock_low
) %>%
dplyr::select(diff_ih_il, diff_ih_mh, diff_ih_ml,
diff_il_mh, diff_il_ml, diff_mh_ml)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
beta_to_mu <- pred_grid_red_2_1_4
beta_to_mu[,4] <- 0
beta_to_mu <- unique(beta_to_mu)
cmat_red <- cmat %*% beta_to_mu
beta_to_mu_full <- pred_grid_full_2_1_4
beta_to_mu_full[,4] <- 0
beta_to_mu_full <- unique(beta_to_mu_full)
cmat_full <- cmat %*% beta_to_mu_full
test_values <- get_test_values_twofactor(data_2_1_4, Xmatrix_2_1_4, Xmatrix_2_1_4_full, cmat_red, cmat_full, cobra$which_X)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3", "C4", "C5", "C6")
test_stat_2_1_4 <- diffs_2_1_4 %>%
cbind(group_counts_2_1_4) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_ih_il = (diff_ih_il / C1),
stat_ih_mh = (diff_ih_mh / C2),
stat_ih_ml = (diff_ih_ml / C3),
stat_il_mh = (diff_il_mh / C4),
stat_il_ml = (diff_il_ml / C5),
stat_mh_ml = (diff_mh_ml / C6)
) %>%
dplyr::select(stat_ih_il, stat_ih_mh, stat_ih_ml,
stat_il_mh, stat_il_ml, stat_mh_ml) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_2_1_4 <- test_stat_2_1_4 %>%
dplyr::mutate(
lg = group_counts_2_1_4$n_grp,
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_ih_il),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_high = pt(
q = abs(stat_ih_mh),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_low = pt(
q = abs(stat_ih_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_high = pt(
q = abs(stat_il_mh),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_low = pt(
q = abs(stat_il_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Mock_high_vs_Mock_low = pt(
q = abs(stat_mh_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low) %>%
`row.names<-`(NULL)
flag_a_2_1_4 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,2],
pvals = pval_a_2_1_4[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_high = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,3],
pvals = pval_a_2_1_4[, 2],
cutoff = 0.05
),
Infection_high_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 3],
cutoff = 0.05
),
Infection_low_vs_Mock_high = aflag(
grp1 = mean_a_2_1_4[,2],
grp2 = mean_a_2_1_4[,3],
pvals = pval_a_2_1_4[, 4],
cutoff = 0.05
),
Infection_low_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,2],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 5],
cutoff = 0.05
),
Mock_high_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,3],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 6],
cutoff = 0.05
)
)
astan_2_1_4 <- data.frame(
Mass_Tag_ID = afilta_2_1_4$e_data$Mass_Tag_ID,
Count_Infection_high = group_counts_2_1_4$nona_Infection_high,
Count_Infection_low = group_counts_2_1_4$nona_Infection_low,
Count_Mock_high = group_counts_2_1_4$nona_Mock_high,
Count_Mock_low = group_counts_2_1_4$nona_Mock_low,
mean_a_2_1_4,
Fold_change_Infection_high_vs_Infection_low = diffs_2_1_4$diff_ih_il,
Fold_change_Infection_high_vs_Mock_high = diffs_2_1_4$diff_ih_mh,
Fold_change_Infection_high_vs_Mock_low = diffs_2_1_4$diff_ih_ml,
Fold_change_Infection_low_vs_Mock_high = diffs_2_1_4$diff_il_mh,
Fold_change_Infection_low_vs_Mock_low = diffs_2_1_4$diff_il_ml,
Fold_change_Mock_high_vs_Mock_low = diffs_2_1_4$diff_mh_ml,
pval_a_2_1_4,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_1_4[, 1],
Flag_A_Infection_high_vs_Mock_high = flag_a_2_1_4[, 2],
Flag_A_Infection_high_vs_Mock_low = flag_a_2_1_4[, 3],
Flag_A_Infection_low_vs_Mock_high = flag_a_2_1_4[, 4],
Flag_A_Infection_low_vs_Mock_low = flag_a_2_1_4[, 5],
Flag_A_Mock_high_vs_Mock_low = flag_a_2_1_4[, 6],
row.names = NULL
)
class(astan_2_1_4) <- c("statRes", "data.frame")
attr(astan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(astan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(astan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_total = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_anova = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_gtest = c(0, 0, 0, 0, 0, 0),
Down_gtest = c(0, 0, 0, 0, 0, 0),
row.names = NULL
)
attr(astan_2_1_4, "statistical_test") <- "anova"
attr(astan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(astan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(astan_2_1_4, "pval_thresh") <- 0.05
attr(astan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_1_4$e_data[, -1])) /
prod(dim(afilta_2_1_4$e_data[, -1]))),
num_samps = dim(afilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(astan_2_1_4, "which_X") <- which_X_2_1_4_a
attr(astan_2_1_4, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_1_4$e_data$Mass_Tag_ID,
flag_a_2_1_4
)
attr(astan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_1_4, "data_class") <- "pepData"
tukey_stat_2_1_4 <- test_stat_2_1_4 * sqrt(2)
n_tests_2_1_4 <- rowSums(!is.na(tukey_stat_2_1_4))
suppressWarnings(
tukey_pval_2_1_4 <- tukey_stat_2_1_4 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_ih_il),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_high = ptukey(
abs(stat_ih_mh),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_low = ptukey(
abs(stat_ih_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_high = ptukey(
abs(stat_il_mh),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_low = ptukey(
abs(stat_il_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Mock_high_vs_Mock_low = ptukey(
abs(stat_mh_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low,) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_1_4 <- which(rowSums(!is.na(tukey_stat_2_1_4)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_1_4 <- which(!is.na(pval_a_2_1_4[just_one_2_1_4, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_1_4[just_one_2_1_4, ][not_na_2_1_4] <- data.matrix(
pval_a_2_1_4[just_one_2_1_4, ]
)[not_na_2_1_4]
# Convert back to a data frame.
tukey_pval_2_1_4 <- data.frame(tukey_pval_2_1_4)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
dunnett_2_1_4 <- pval_a_2_1_4 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_1_4 - no_nas) %>%
cbind(test_stat_2_1_4) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_ih_il = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_il),
no_nas),
rep(abs(stat_ih_il),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_mh),
no_nas),
rep(abs(stat_ih_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_ml),
no_nas),
rep(abs(stat_ih_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_mh),
no_nas),
rep(abs(stat_il_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_ml),
no_nas),
rep(abs(stat_il_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_mh_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_mh_ml),
no_nas),
rep(abs(stat_mh_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_ih_il, pval_ih_mh, pval_ih_ml,
pval_il_mh, pval_il_ml, pval_mh_ml)
# main effects: 2; covariates: 2; groups: 4 ---------------
data_2_2_4 = data.matrix(afilta_2_2_4$e_data[, -1])
covariate_names = colnames(attr(attr(afilta_2_2_4, "group_DF"), "covariates"))[-1]
main_effect_names = attr(attr(afilta_2_2_4, "group_DF"), "main_effects")
group_sampnames <- groupDF_2_2_4[,get_fdata_cname(afilta_2_2_4)]
groupData <- groupDF_2_2_4[group_sampnames %in% colnames(afilta_2_2_4$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_2_2_4$f_data)
groupData[,main_effect_names] <- lapply(groupData[main_effect_names], function(x) factor(x, levels=unique(x)))
pred_grid_red_2_2_4 = get_pred_grid(groupData, main_effect_names, covariate_names)
pred_grid_full_2_2_4 = get_pred_grid(groupData, main_effect_names, covariate_names, fspec = as.formula("~Condition*Level+Gender+Age"))
cobra <- run_twofactor_cpp(
data = data.matrix(data_2_2_4),
gpData = groupData[,c("Group", main_effect_names, covariate_names)],
Xfull=Xmatrix_2_2_4_full, Xred = Xmatrix_2_2_4,
pred_grid_full = pred_grid_full_2_2_4, pred_grid_red = pred_grid_red_2_2_4,
continuous_covar_inds = 5
)
group_counts_2_2_4 <- data.frame(
nona_Infection_high = rowSums(!is.na(data_2_2_4[, c(1, 3, 5)])),
nona_Infection_low = rowSums(!is.na(data_2_2_4[, c(2, 4, 6)])),
nona_Mock_high = rowSums(!is.na(data_2_2_4[, c(7, 8, 12)])),
nona_Mock_low = rowSums(!is.na(data_2_2_4[, c(9, 10, 11)]))
) %>%
dplyr::ungroup()
nona_grps_2_2_4 <- unname(rowSums(group_counts_2_2_4 != 0))
nona_counts_2_2_4 <- unname(rowSums(!is.na(data_2_2_4)))
mean_a_2_2_4 <- data.frame(cobra$lsmeans)
colnames(mean_a_2_2_4) <- paste0("Mean_", unique(groupData$Group))
diffs_2_2_4 <- mean_a_2_2_4 %>%
dplyr::mutate(
diff_ih_il = Mean_Infection_high - Mean_Infection_low,
diff_ih_mh = Mean_Infection_high - Mean_Mock_high,
diff_ih_ml = Mean_Infection_high - Mean_Mock_low,
diff_il_mh = Mean_Infection_low - Mean_Mock_high,
diff_il_ml = Mean_Infection_low - Mean_Mock_low,
diff_mh_ml = Mean_Mock_high - Mean_Mock_low
) %>%
dplyr::select(diff_ih_il, diff_ih_mh, diff_ih_ml,
diff_il_mh, diff_il_ml, diff_mh_ml)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
beta_to_mu = pred_grid_red_2_2_4
beta_to_mu[,4:5] <- 0
beta_to_mu <- unique(beta_to_mu)
cmat_red <- cmat %*% beta_to_mu
beta_to_mu_full = pred_grid_full_2_2_4
beta_to_mu_full[,4:5] <- 0
beta_to_mu_full <- unique(beta_to_mu_full)
cmat_full <- cmat %*% beta_to_mu_full
test_values <- get_test_values_twofactor(data_2_2_4, Xmatrix_2_2_4, Xmatrix_2_2_4_full, cmat_red, cmat_full, cobra$which_X)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3", "C4", "C5", "C6")
test_stat_2_2_4 <- diffs_2_2_4 %>%
cbind(group_counts_2_2_4) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_ih_il = (diff_ih_il / C1),
stat_ih_mh = (diff_ih_mh / C2),
stat_ih_ml = (diff_ih_ml / C3),
stat_il_mh = (diff_il_mh / C4),
stat_il_ml = (diff_il_ml / C5),
stat_mh_ml = (diff_mh_ml / C6)
) %>%
dplyr::select(stat_ih_il, stat_ih_mh, stat_ih_ml,
stat_il_mh, stat_il_ml, stat_mh_ml) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_2_2_4 <- test_stat_2_2_4 %>%
dplyr::mutate(
lg = group_counts_2_2_4$n_grp,
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_ih_il),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_high = pt(
q = abs(stat_ih_mh),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_low = pt(
q = abs(stat_ih_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_high = pt(
q = abs(stat_il_mh),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_low = pt(
q = abs(stat_il_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Mock_high_vs_Mock_low = pt(
q = abs(stat_mh_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low) %>%
`row.names<-`(NULL)
flag_a_2_2_4 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,2],
pvals = pval_a_2_2_4[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_high = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,3],
pvals = pval_a_2_2_4[, 2],
cutoff = 0.05
),
Infection_high_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 3],
cutoff = 0.05
),
Infection_low_vs_Mock_high = aflag(
grp1 = mean_a_2_2_4[,2],
grp2 = mean_a_2_2_4[,3],
pvals = pval_a_2_2_4[, 4],
cutoff = 0.05
),
Infection_low_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,2],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 5],
cutoff = 0.05
),
Mock_high_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,3],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 6],
cutoff = 0.05
)
)
astan_2_2_4 <- data.frame(
Mass_Tag_ID = afilta_2_2_4$e_data$Mass_Tag_ID,
Count_Infection_high = group_counts_2_2_4$nona_Infection_high,
Count_Infection_low = group_counts_2_2_4$nona_Infection_low,
Count_Mock_high = group_counts_2_2_4$nona_Mock_high,
Count_Mock_low = group_counts_2_2_4$nona_Mock_low,
mean_a_2_2_4,
Fold_change_Infection_high_vs_Infection_low = diffs_2_2_4$diff_ih_il,
Fold_change_Infection_high_vs_Mock_high = diffs_2_2_4$diff_ih_mh,
Fold_change_Infection_high_vs_Mock_low = diffs_2_2_4$diff_ih_ml,
Fold_change_Infection_low_vs_Mock_high = diffs_2_2_4$diff_il_mh,
Fold_change_Infection_low_vs_Mock_low = diffs_2_2_4$diff_il_ml,
Fold_change_Mock_high_vs_Mock_low = diffs_2_2_4$diff_mh_ml,
pval_a_2_2_4,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_2_4[, 1],
Flag_A_Infection_high_vs_Mock_high = flag_a_2_2_4[, 2],
Flag_A_Infection_high_vs_Mock_low = flag_a_2_2_4[, 3],
Flag_A_Infection_low_vs_Mock_high = flag_a_2_2_4[, 4],
Flag_A_Infection_low_vs_Mock_low = flag_a_2_2_4[, 5],
Flag_A_Mock_high_vs_Mock_low = flag_a_2_2_4[, 6],
row.names = NULL
)
class(astan_2_2_4) <- c("statRes", "data.frame")
attr(astan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(astan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(astan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_total = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_anova = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_gtest = c(0, 0, 0, 0, 0, 0),
Down_gtest = c(0, 0, 0, 0, 0, 0),
row.names = NULL
)
attr(astan_2_2_4, "statistical_test") <- "anova"
attr(astan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(astan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(astan_2_2_4, "pval_thresh") <- 0.05
attr(astan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_2_4$e_data[, -1])) /
prod(dim(afilta_2_2_4$e_data[, -1]))),
num_samps = dim(afilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(astan_2_2_4, "which_X") <- which_X_2_2_4_a
attr(astan_2_2_4, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_2_4$e_data$Mass_Tag_ID,
flag_a_2_2_4
)
attr(astan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_2_4, "data_class") <- "pepData"
tukey_stat_2_2_4 <- test_stat_2_2_4 * sqrt(2)
n_tests_2_2_4 <- rowSums(!is.na(tukey_stat_2_2_4))
suppressWarnings(
tukey_pval_2_2_4 <- tukey_stat_2_2_4 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_ih_il),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_high = ptukey(
abs(stat_ih_mh),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_low = ptukey(
abs(stat_ih_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_high = ptukey(
abs(stat_il_mh),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_low = ptukey(
abs(stat_il_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Mock_high_vs_Mock_low = ptukey(
abs(stat_mh_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low,) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_2_4 <- which(rowSums(!is.na(tukey_stat_2_2_4)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_2_4 <- which(!is.na(pval_a_2_2_4[just_one_2_2_4, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_2_4[just_one_2_2_4, ][not_na_2_2_4] <- data.matrix(
pval_a_2_2_4[just_one_2_2_4, ]
)[not_na_2_2_4]
# Convert back to a data frame.
tukey_pval_2_2_4 <- data.frame(tukey_pval_2_2_4)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
suppressWarnings(
dunnett_2_2_4 <- pval_a_2_2_4 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_2_4 - no_nas) %>%
cbind(test_stat_2_2_4) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_ih_il = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_il),
no_nas),
rep(abs(stat_ih_il),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_mh),
no_nas),
rep(abs(stat_ih_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_ml),
no_nas),
rep(abs(stat_ih_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_mh),
no_nas),
rep(abs(stat_il_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_ml),
no_nas),
rep(abs(stat_il_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_mh_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_mh_ml),
no_nas),
rep(abs(stat_mh_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_ih_il, pval_ih_mh, pval_ih_ml,
pval_il_mh, pval_il_ml, pval_mh_ml)
)
# Generate G-Test standards ----------------------------------------------------
# Construct count matrices: Objects that start with obs represent counts of
# non-missing (observed) values and objects that start with abs represent
# counts of missing (absent) values.
obs_inf_1_0_2 <- rowSums(!is.na(gfilta_1_0_2$e_data[, 2:10]))
obs_mock_1_0_2 <- rowSums(!is.na(gfilta_1_0_2$e_data[, 11:13]))
abs_inf_1_0_2 <- rowSums(is.na(gfilta_1_0_2$e_data[, 2:10]))
abs_mock_1_0_2 <- rowSums(is.na(gfilta_1_0_2$e_data[, 11:13]))
obs_mut_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(2, 4, 5, 10)]))
obs_zom_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(3, 6:9)]))
obs_hum_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(11:13)]))
abs_mut_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(2, 4, 5, 10)]))
abs_zom_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(3, 6:9)]))
abs_hum_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(11:13)]))
obs_mut_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(2, 4, 5, 10)]))
obs_zom_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(3, 6:9)]))
obs_hum_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(11:13)]))
abs_mut_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(2, 4, 5, 10)]))
abs_zom_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(3, 6:9)]))
abs_hum_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(11:13)]))
obs_inf_high_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, c(2:4, 7:8)]))
obs_inf_low_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, c(5:6, 9:10)]))
obs_mock_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, 11:13]))
abs_inf_high_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, c(2:4, 7:8)]))
abs_inf_low_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, c(5:6, 9:10)]))
abs_mock_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, 11:13]))
obs_ih_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(2, 4, 6)]))
obs_il_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(3, 5, 7)]))
obs_mh_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(8, 9, 13)]))
obs_ml_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, 10:12]))
abs_ih_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(2, 4, 6)]))
abs_il_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(3, 5, 7)]))
abs_mh_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(8, 9, 13)]))
abs_ml_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, 10:12]))
obs_ih_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(2, 4, 6)]))
obs_il_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(3, 5, 7)]))
obs_mh_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(8, 9, 13)]))
obs_ml_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, 10:12]))
abs_ih_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(2, 4, 6)]))
abs_il_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(3, 5, 7)]))
abs_mh_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(8, 9, 13)]))
abs_ml_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, 10:12]))
# main effects: 1; covariates: 0; groups: 2 ---------------
pval_g_1_0_2 <- data.frame(
Infection_vs_Mock = pchisq(
q = g(obs1 = obs_inf_1_0_2,
obs2 = obs_mock_1_0_2,
abs1 = abs_inf_1_0_2,
abs2 = abs_mock_1_0_2,
n1 = 9,
n2 = 3,
n_total = 9 + 3),
df = 1,
lower.tail = FALSE
)
)
flag_g_1_0_2 <- data.frame(
Infection_vs_Mock = gflag(
obs1 = obs_inf_1_0_2,
obs2 = obs_mock_1_0_2,
abs1 = abs_inf_1_0_2,
abs2 = abs_mock_1_0_2,
pvals = pval_g_1_0_2[, 1],
cutoff = 0.05
)
)
mean_1_0_2 <- data.frame(
Mean_Infection = rowMeans(gfilta_1_0_2$e_data[, 2:10],
na.rm = TRUE),
Mean_Mock = rowMeans(gfilta_1_0_2$e_data[, 11:13],
na.rm = TRUE)
)
gstan_1_0_2 <- data.frame(
Mass_Tag_ID = gfilta_1_0_2$e_data$Mass_Tag_ID,
Count_Infection = unname(obs_inf_1_0_2),
Count_Mock = unname(obs_mock_1_0_2),
mean_1_0_2,
Fold_change_Infection_vs_Mock = (
mean_1_0_2[, 1] - mean_1_0_2[, 2]
),
P_value_G_Infection_vs_Mock = pval_g_1_0_2[, 1],
Flag_G_Infection_vs_Mock = flag_g_1_0_2[, 1],
row.names = NULL
)
class(gstan_1_0_2) <- c("statRes", "data.frame")
attr(gstan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(gstan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(gstan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_total = c(length(which(flag_g_1_0_2[, 1] == -1))),
Up_anova = c(0),
Down_anova = c(0),
Up_gtest = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_gtest = c(length(which(flag_g_1_0_2[, 1] == -1))),
row.names = NULL
)
attr(gstan_1_0_2, "statistical_test") <- "gtest"
attr(gstan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_0_2, "pval_thresh") <- 0.05
attr(gstan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_0_2$e_data[, -1])) /
prod(dim(gfilta_1_0_2$e_data[, -1]))),
num_samps = dim(gfilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(gstan_1_0_2, "which_X") <- rep(0, nrow(gstan_1_0_2))
attr(gstan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
pval_g_1_1_3 <- data.frame(
P_value_G_mutant_vs_zombie = rep(0, nrow(gfilta_1_1_3$e_data)),
P_value_G_mutant_vs_human = rep(0, nrow(gfilta_1_1_3$e_data)),
P_value_G_zombie_vs_human = rep(0, nrow(gfilta_1_1_3$e_data))
)
for (e in 1:nrow(gfilta_1_1_3$e_data)) {
pval_g_1_1_3[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, 2:10])) ~
attr(groupDF_1_1_3, "covariates")$Gender[1:9] +
groupDF_1_1_3$Group[1:9],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
pval_g_1_1_3[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, c(2, 4, 5, 10:13)])) ~
attr(groupDF_1_1_3, "covariates")$Gender[c(1, 3, 4, 9:12)] +
groupDF_1_1_3$Group[c(1, 3, 4, 9:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
pval_g_1_1_3[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, c(3, 6:9, 11:13)])) ~
attr(groupDF_1_1_3, "covariates")$Gender[c(2, 5:8, 10:12)] +
groupDF_1_1_3$Group[c(2, 5:8, 10:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
}
flag_g_1_1_3 <- data.frame(
Flag_G_mutant_vs_zombie = gflag(
obs1 = obs_mut_1_1_3,
obs2 = obs_zom_1_1_3,
abs1 = abs_mut_1_1_3,
abs2 = abs_zom_1_1_3,
pvals = pval_g_1_1_3[, 1],
cutoff = 0.05
),
Flag_G_mutant_vs_human = gflag(
obs1 = obs_mut_1_1_3,
obs2 = obs_hum_1_1_3,
abs1 = abs_mut_1_1_3,
abs2 = abs_hum_1_1_3,
pvals = pval_g_1_1_3[, 2],
cutoff = 0.05
),
Flag_G_zombie_vs_human = gflag(
obs1 = obs_zom_1_1_3,
obs2 = obs_hum_1_1_3,
abs1 = abs_zom_1_1_3,
abs2 = abs_hum_1_1_3,
pvals = pval_g_1_1_3[, 3],
cutoff = 0.05
)
)
data_1_1_3 <- gfilta_1_1_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_1_3), Xmatrix = data.matrix(Xmatrix_1_1_3))
group_sampnames <- gfilta_1_1_3$f_data$SampleID
groupData <- groupDF_1_1_3[group_sampnames %in% colnames(gfilta_1_1_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(gfilta_1_1_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", "Gender")
mean_1_1_3 <- get_lsmeans(data = data_1_1_3, xmatrix = Xmatrix_1_1_3, pred_grid = pred_grid, Betas = Betas)
counts_1_1_3 <- data.frame(
"Count_mutant" = unname(obs_mut_1_1_3),
"Count_zombie" = unname(obs_zom_1_1_3),
"Count_human" = unname(obs_hum_1_1_3)
)
mean_1_1_3[counts_1_1_3 == 0] <- NA
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
diffs_1_1_3 <- fold_change_diff(data.matrix(mean_1_1_3), cmat)
gstan_1_1_3 <- data.frame(
Mass_Tag_ID = gfilta_1_1_3$e_data$Mass_Tag_ID,
counts_1_1_3,
mean_1_1_3,
Fold_change_mutant_vs_zombie = diffs_1_1_3[,1],
Fold_change_mutant_vs_human = diffs_1_1_3[,2],
Fold_change_zombie_vs_human = diffs_1_1_3[,3],
pval_g_1_1_3,
flag_g_1_1_3,
row.names = NULL
)
class(gstan_1_1_3) <- c("statRes", "data.frame")
attr(gstan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(gstan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(gstan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_total = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_1_1_3, "statistical_test") <- "gtest"
attr(gstan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_1_3, "pval_thresh") <- 0.05
attr(gstan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_1_3$e_data[, -1])) /
prod(dim(gfilta_1_1_3$e_data[, -1]))),
num_samps = dim(gfilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(gstan_1_1_3, "which_X") <- rep(0, nrow(gstan_1_1_3))
attr(gstan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_1_3, "data_class") <- "pepData"
# main effects: 1; covariates: 2; groups: 3 ---------------
pval_g_1_2_3 <- data.frame(
P_value_G_mutant_vs_zombie = rep(0, nrow(gfilta_1_2_3$e_data)),
P_value_G_mutant_vs_human = rep(0, nrow(gfilta_1_2_3$e_data)),
P_value_G_zombie_vs_human = rep(0, nrow(gfilta_1_2_3$e_data))
)
for (e in 1:nrow(gfilta_1_2_3$e_data)) {
pval_g_1_2_3[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, 2:10])) ~
attr(groupDF_1_2_3, "covariates")$Gender[1:9] +
attr(groupDF_1_2_3, "covariates")$Age[1:9] +
groupDF_1_2_3$Group[1:9],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
pval_g_1_2_3[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, c(2, 4, 5, 10:13)])) ~
attr(groupDF_1_2_3, "covariates")$Gender[c(1, 3, 4, 9:12)] +
attr(groupDF_1_2_3, "covariates")$Age[c(1, 3, 4, 9:12)] +
groupDF_1_2_3$Group[c(1, 3, 4, 9:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
pval_g_1_2_3[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, c(3, 6:9, 11:13)])) ~
attr(groupDF_1_2_3, "covariates")$Gender[c(2, 5:8, 10:12)] +
attr(groupDF_1_2_3, "covariates")$Age[c(2, 5:8, 10:12)] +
groupDF_1_2_3$Group[c(2, 5:8, 10:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
}
flag_g_1_2_3 <- data.frame(
Flag_G_mutant_vs_zombie = gflag(
obs1 = obs_mut_1_2_3,
obs2 = obs_zom_1_2_3,
abs1 = abs_mut_1_2_3,
abs2 = abs_zom_1_2_3,
pvals = pval_g_1_2_3[, 1],
cutoff = 0.05
),
Flag_G_mutant_vs_human = gflag(
obs1 = obs_mut_1_2_3,
obs2 = obs_hum_1_2_3,
abs1 = abs_mut_1_2_3,
abs2 = abs_hum_1_2_3,
pvals = pval_g_1_2_3[, 2],
cutoff = 0.05
),
Flag_G_zombie_vs_human = gflag(
obs1 = obs_zom_1_2_3,
obs2 = obs_hum_1_2_3,
abs1 = abs_zom_1_2_3,
abs2 = abs_hum_1_2_3,
pvals = pval_g_1_2_3[, 3],
cutoff = 0.05
)
)
data_1_2_3 = data.matrix(gfilta_1_2_3$e_data[, -1])
Betas = compute_betas(data_mat = data_1_2_3, Xmatrix = data.matrix(Xmatrix_1_2_3))
group_sampnames <- gfilta_1_2_3$f_data$SampleID
groupData <- groupDF_1_2_3[group_sampnames %in% colnames(gfilta_1_2_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(gfilta_1_2_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", c("Gender", "Age"))
mean_1_2_3 <- get_lsmeans(data = data_1_2_3, xmatrix = Xmatrix_1_2_3, pred_grid = pred_grid, Betas = Betas, continuous_covar_inds = 5)
counts_1_2_3 <- data.frame(
"Count_mutant" = unname(obs_mut_1_2_3),
"Count_zombie" = unname(obs_zom_1_2_3),
"Count_human" = unname(obs_hum_1_2_3)
)
mean_1_2_3[counts_1_2_3 == 0] <- NA
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
diffs_1_2_3 <- fold_change_diff(data.matrix(mean_1_2_3), cmat)
gstan_1_2_3 <- data.frame(
Mass_Tag_ID = gfilta_1_2_3$e_data$Mass_Tag_ID,
counts_1_2_3,
mean_1_2_3,
Fold_change_mutant_vs_zombie = diffs_1_2_3[,1],
Fold_change_mutant_vs_human = diffs_1_2_3[,2],
Fold_change_zombie_vs_human = diffs_1_2_3[,3],
pval_g_1_2_3,
flag_g_1_2_3,
row.names = NULL
)
class(gstan_1_2_3) <- c("statRes", "data.frame")
attr(gstan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(gstan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(gstan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_total = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_1_2_3, "statistical_test") <- "gtest"
attr(gstan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_2_3, "pval_thresh") <- 0.05
attr(gstan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_2_3$e_data[, -1])) /
prod(dim(gfilta_1_2_3$e_data[, -1]))),
num_samps = dim(gfilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(gstan_1_2_3, "which_X") <- rep(0, nrow(gstan_1_2_3))
attr(gstan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_2_3, "data_class") <- "pepData"
# main effects: 2; covariates: 0; groups: 3 ---------------
pval_g_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = pchisq(
q = g(obs1 = obs_inf_high_2_0_3,
obs2 = obs_inf_low_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_inf_low_2_0_3,
n1 = 5,
n2 = 4,
n_total = 5 + 4),
df = 1,
lower.tail = FALSE
),
Infection_high_vs_Mock_none = pchisq(
q = g(obs1 = obs_inf_high_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_mock_2_0_3,
n1 = 5,
n2 = 3,
n_total = 5 + 3),
df = 1,
lower.tail = FALSE
),
Infection_low_vs_Mock_none = pchisq(
q = g(obs1 = obs_inf_low_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_low_2_0_3,
abs2 = abs_mock_2_0_3,
n1 = 4,
n2 = 3,
n_total = 4 + 3),
df = 1,
lower.tail = FALSE
)
)
flag_g_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = gflag(
obs1 = obs_inf_high_2_0_3,
obs2 = obs_inf_low_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_inf_low_2_0_3,
pvals = pval_g_2_0_3[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_none = gflag(
obs1 = obs_inf_high_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_mock_2_0_3,
pvals = pval_g_2_0_3[, 2],
cutoff = 0.05
),
Infection_low_vs_Mock_none = gflag(
obs1 = obs_inf_low_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_low_2_0_3,
abs2 = abs_mock_2_0_3,
pvals = pval_g_2_0_3[, 3],
cutoff = 0.05
)
)
mean_2_0_3 <- data.frame(
Mean_Infection_high = rowMeans(gfilta_2_0_3$e_data[, c(2:4, 7:8)],
na.rm = TRUE),
Mean_Infection_low = rowMeans(gfilta_2_0_3$e_data[, c(5:6, 9:10)],
na.rm = TRUE),
Mean_Mock_none = rowMeans(gfilta_2_0_3$e_data[, 11:13],
na.rm = TRUE)
)
gstan_2_0_3 <- data.frame(
Mass_Tag_ID = gfilta_2_0_3$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_inf_high_2_0_3),
Count_Infection_low = unname(obs_inf_low_2_0_3),
Count_Mock_none = unname(obs_mock_2_0_3),
mean_2_0_3,
Fold_change_Infection_high_vs_Infection_low = (
mean_2_0_3[, 1] - mean_2_0_3[, 2]
),
Fold_change_Infection_high_vs_Mock_none = (
mean_2_0_3[, 1] - mean_2_0_3[, 3]
),
Fold_change_Infection_low_vs_Mock_none = (
mean_2_0_3[, 2] - mean_2_0_3[, 3]
),
P_value_G_Infection_high_vs_Infection_low = pval_g_2_0_3[, 1],
P_value_G_Infection_high_vs_Mock_none = pval_g_2_0_3[, 2],
P_value_G_Infection_low_vs_Mock_none = pval_g_2_0_3[, 3],
Flag_G_Infection_high_vs_Infection_low = flag_g_2_0_3[, 1],
Flag_G_Infection_high_vs_Mock_none = flag_g_2_0_3[, 2],
Flag_G_Infection_low_vs_Mock_none = flag_g_2_0_3[, 3],
row.names = NULL
)
class(gstan_2_0_3) <- c("statRes", "data.frame")
attr(gstan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(gstan_2_0_3, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none")
attr(gstan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_total = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_2_0_3, "statistical_test") <- "gtest"
attr(gstan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_0_3, "pval_thresh") <- 0.05
attr(gstan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_0_3$e_data[, -1])) /
prod(dim(gfilta_2_0_3$e_data[, -1]))),
num_samps = dim(gfilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(gstan_2_0_3, "which_X") <- rep(0, nrow(gstan_2_0_3))
attr(gstan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_0_3, "data_class") <- "pepData"
# main effects: 2; covariates: 1; groups: 4 ---------------
pval_g_2_1_4 <- data.frame(
P_value_G_Infection_high_vs_Infection_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_high_vs_Mock_high = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_high_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_low_vs_Mock_high = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_low_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Mock_high_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data))
)
for (e in 1:nrow(gfilta_2_1_4$e_data)) {
# Compute the p-value for the Infection_high_vs_Infection_low test.
pval_g_2_1_4[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, 2:7])) ~
attr(groupDF_2_1_4, "covariates")$Gender[1:6] +
groupDF_2_1_4$Group[1:6],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_high_vs_Mock_high test.
pval_g_2_1_4[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(2, 4, 6, 8, 9, 13)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(1, 3, 5, 7, 8, 12)] +
groupDF_2_1_4$Group[c(1, 3, 5, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_high_vs_Mock_low test.
pval_g_2_1_4[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(2, 4, 6, 10:12)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(1, 3, 5, 9:11)] +
groupDF_2_1_4$Group[c(1, 3, 5, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_low_vs_Mock_high test.
pval_g_2_1_4[e, 4] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(3, 5, 7, 8, 9, 13)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(2, 4, 6, 7, 8, 12)] +
groupDF_2_1_4$Group[c(2, 4, 6, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_low_vs_Mock_low test.
pval_g_2_1_4[e, 5] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(3, 5, 7, 10:12)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(2, 4, 6, 9:11)] +
groupDF_2_1_4$Group[c(2, 4, 6, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Mock_high_vs_Mock_low test.
pval_g_2_1_4[e, 6] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, 8:13])) ~
attr(groupDF_2_1_4, "covariates")$Gender[7:12] +
groupDF_2_1_4$Group[7:12],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
}
flag_g_2_1_4 <- data.frame(
Flag_G_Infection_high_vs_Infection_low = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_il_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_il_2_1_4,
pvals = pval_g_2_1_4[, 1],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_high = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_mh_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_mh_2_1_4,
pvals = pval_g_2_1_4[, 2],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_low = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 3],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_high = gflag(
obs1 = obs_il_2_1_4,
obs2 = obs_mh_2_1_4,
abs1 = abs_il_2_1_4,
abs2 = abs_mh_2_1_4,
pvals = pval_g_2_1_4[, 4],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_low = gflag(
obs1 = obs_il_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_il_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 5],
cutoff = 0.05
),
Flag_G_Mock_high_vs_Mock_low = gflag(
obs1 = obs_mh_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_mh_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 6],
cutoff = 0.05
)
)
gdf = dplyr::left_join(groupDF_2_1_4, attr(groupDF_2_1_4, "covariates")) %>%
dplyr::select(-SampleID)
dragon <- run_twofactor_cpp(
data = data.matrix(gfilta_2_1_4$e_data[, -1]),
gpData = gdf,
Xfull = Xmatrix_2_1_4_full,
Xred = Xmatrix_2_1_4,
pred_grid_full = pred_grid_full_2_1_4,
pred_grid_red = pred_grid_red_2_1_4,
continuous_covar_inds = numeric(0)
)
mean_2_1_4 <- data.frame(
Mean_Infection_high = dragon$adj_group_means[, 1],
Mean_Infection_low = dragon$adj_group_means[, 2],
Mean_Mock_high = dragon$adj_group_means[, 3],
Mean_Mock_low = dragon$adj_group_means[, 4]
)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
diffs_2_1_4 <- fold_change_diff(data.matrix(mean_2_1_4), cmat)
diffs_2_1_4 <- as.data.frame(diffs_2_1_4) %>%
`colnames<-`(c("Fold_change_Infection_high_vs_Infection_low", "Fold_change_Infection_high_vs_Mock_high", "Fold_change_Infection_high_vs_Mock_low", "Fold_change_Infection_low_vs_Mock_high", "Fold_change_Infection_low_vs_Mock_low", "Fold_change_Mock_high_vs_Mock_low"))
gstan_2_1_4 <- data.frame(
Mass_Tag_ID = gfilta_2_1_4$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_ih_2_1_4),
Count_Infection_low = unname(obs_il_2_1_4),
Count_Mock_high = unname(obs_mh_2_1_4),
Count_Mock_low = unname(obs_ml_2_1_4),
mean_2_1_4,
diffs_2_1_4,
pval_g_2_1_4,
flag_g_2_1_4,
row.names = NULL
)
class(gstan_2_1_4) <- c("statRes", "data.frame")
attr(gstan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(gstan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(gstan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_total = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
Up_anova = c(0, 0, 0, 0, 0, 0),
Down_anova = c(0, 0, 0, 0, 0, 0),
Up_gtest = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
row.names = NULL
)
attr(gstan_2_1_4, "statistical_test") <- "gtest"
attr(gstan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_1_4, "pval_thresh") <- 0.05
attr(gstan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_1_4$e_data[, -1])) /
prod(dim(gfilta_2_1_4$e_data[, -1]))),
num_samps = dim(gfilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(gstan_2_1_4, "which_X") <- which_X_2_1_4_g
attr(gstan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_1_4, "data_class") <- "pepData"
# main effects: 2; covariates: 2; groups: 4 ---------------
pval_g_2_2_4 <- data.frame(
P_value_G_Infection_high_vs_Infection_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_high_vs_Mock_high = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_high_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_low_vs_Mock_high = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_low_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Mock_high_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data))
)
for (e in 1:nrow(gfilta_2_2_4$e_data)) {
# Compute the p-value for the Infection_high_vs_Infection_low test.
pval_g_2_2_4[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, 2:7])) ~
attr(groupDF_2_2_4, "covariates")$Gender[1:6] +
attr(groupDF_2_2_4, "covariates")$Age[1:6] +
groupDF_2_2_4$Group[1:6],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_high_vs_Mock_high test.
pval_g_2_2_4[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(2, 4, 6, 8, 9, 13)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(1, 3, 5, 7, 8, 12)] +
attr(groupDF_2_2_4, "covariates")$Age[c(1, 3, 5, 7, 8, 12)] +
groupDF_2_2_4$Group[c(1, 3, 5, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_high_vs_Mock_low test.
pval_g_2_2_4[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(2, 4, 6, 10:12)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(1, 3, 5, 9:11)] +
attr(groupDF_2_2_4, "covariates")$Age[c(1, 3, 5, 9:11)] +
groupDF_2_2_4$Group[c(1, 3, 5, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_low_vs_Mock_high test.
pval_g_2_2_4[e, 4] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(3, 5, 7, 8, 9, 13)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(2, 4, 6, 7, 8, 12)] +
attr(groupDF_2_2_4, "covariates")$Age[c(2, 4, 6, 7, 8, 12)] +
groupDF_2_2_4$Group[c(2, 4, 6, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_low_vs_Mock_low test.
pval_g_2_2_4[e, 5] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(3, 5, 7, 10:12)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(2, 4, 6, 9:11)] +
attr(groupDF_2_2_4, "covariates")$Age[c(2, 4, 6, 9:11)] +
groupDF_2_2_4$Group[c(2, 4, 6, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Mock_high_vs_Mock_low test.
pval_g_2_2_4[e, 6] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, 8:13])) ~
attr(groupDF_2_2_4, "covariates")$Gender[7:12] +
attr(groupDF_2_2_4, "covariates")$Age[7:12] +
groupDF_2_2_4$Group[7:12],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
}
flag_g_2_2_4 <- data.frame(
Flag_G_Infection_high_vs_Infection_low = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_il_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_il_2_2_4,
pvals = pval_g_2_2_4[, 1],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_high = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_mh_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_mh_2_2_4,
pvals = pval_g_2_2_4[, 2],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_low = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 3],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_high = gflag(
obs1 = obs_il_2_2_4,
obs2 = obs_mh_2_2_4,
abs1 = abs_il_2_2_4,
abs2 = abs_mh_2_2_4,
pvals = pval_g_2_2_4[, 4],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_low = gflag(
obs1 = obs_il_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_il_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 5],
cutoff = 0.05
),
Flag_G_Mock_high_vs_Mock_low = gflag(
obs1 = obs_mh_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_mh_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 6],
cutoff = 0.05
)
)
gdf = dplyr::left_join(groupDF_2_2_4, attr(groupDF_2_2_4, "covariates")) %>%
dplyr::select(-SampleID)
mustang <- run_twofactor_cpp(
data = data.matrix(gfilta_2_2_4$e_data[, -1]),
gpData = gdf,
Xfull=Xmatrix_2_2_4_full, Xred = Xmatrix_2_2_4,
pred_grid_full = pred_grid_full_2_2_4, pred_grid_red = pred_grid_red_2_2_4,
continuous_covar_inds = 5
)
mean_2_2_4 <- data.frame(
Mean_Infection_high = mustang$adj_group_means[, 1],
Mean_Infection_low = mustang$adj_group_means[, 2],
Mean_Mock_high = mustang$adj_group_means[, 3],
Mean_Mock_low = mustang$adj_group_means[, 4]
)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
diffs_2_2_4 <- fold_change_diff(data.matrix(mean_2_2_4), cmat)
diffs_2_2_4 <- as.data.frame(diffs_2_2_4) %>%
`colnames<-`(c("Fold_change_Infection_high_vs_Infection_low", "Fold_change_Infection_high_vs_Mock_high", "Fold_change_Infection_high_vs_Mock_low", "Fold_change_Infection_low_vs_Mock_high", "Fold_change_Infection_low_vs_Mock_low", "Fold_change_Mock_high_vs_Mock_low"))
gstan_2_2_4 <- data.frame(
Mass_Tag_ID = gfilta_2_2_4$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_ih_2_2_4),
Count_Infection_low = unname(obs_il_2_2_4),
Count_Mock_high = unname(obs_mh_2_2_4),
Count_Mock_low = unname(obs_ml_2_2_4),
mean_2_2_4,
diffs_2_2_4,
pval_g_2_2_4,
flag_g_2_2_4,
row.names = NULL
)
class(gstan_2_2_4) <- c("statRes", "data.frame")
attr(gstan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(gstan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(gstan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_total = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
Up_anova = c(0, 0, 0, 0, 0, 0),
Down_anova = c(0, 0, 0, 0, 0, 0),
Up_gtest = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
row.names = NULL
)
attr(gstan_2_2_4, "statistical_test") <- "gtest"
attr(gstan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_2_4, "pval_thresh") <- 0.05
attr(gstan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_2_4$e_data[, -1])) /
prod(dim(gfilta_2_2_4$e_data[, -1]))),
num_samps = dim(gfilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(gstan_2_2_4, "which_X") <- which_X_2_2_4_g
attr(gstan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_2_4, "data_class") <- "pepData"
# Create combined standards ----------------------------------------------------
# main effects: 1; covariates: 0; groups: 2 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_0_2 <- dplyr::full_join(gstan_1_0_2[, c(1:3, 7:8)],
astan_1_0_2) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_0_2[is.nan(data.matrix(cstan_1_0_2))] <- NA
class(cstan_1_0_2) <- c("statRes", "data.frame")
attr(cstan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(cstan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(cstan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = sum(length(which(flag_g_1_0_2[, 1] == 1)),
length(which(flag_a_1_0_2[, 1] == 1))),
Down_total = sum(length(which(flag_g_1_0_2[, 1] == -1)),
length(which(flag_a_1_0_2[, 1] == -1))),
Up_anova = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_anova = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_gtest = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_gtest = c(length(which(flag_g_1_0_2[, 1] == -1))),
row.names = NULL
)
attr(cstan_1_0_2, "statistical_test") <- "combined"
attr(cstan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_0_2, "pval_thresh") <- 0.05
attr(cstan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_0_2$e_data[, -1])) /
prod(dim(gfilta_1_0_2$e_data[, -1]))),
num_samps = dim(gfilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(cstan_1_0_2, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, NA,
0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, NA,
0, 0, NA, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0,
0, 0, NA, NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0,
0, 0, NA, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA,
0, 0)
attr(cstan_1_0_2, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_0_2$e_data$Mass_Tag_ID,
Infection_vs_Mock = dplyr::case_when(
is.na(cstan_1_0_2$Flag_A_Infection_vs_Mock) ~
cstan_1_0_2$Flag_G_Infection_vs_Mock,
(cstan_1_0_2$P_value_A_Infection_vs_Mock > 0.05 &
cstan_1_0_2$P_value_G_Infection_vs_Mock < 0.05) ~
cstan_1_0_2$Flag_G_Infection_vs_Mock,
!is.na(cstan_1_0_2$Flag_A_Infection_vs_Mock) ~
cstan_1_0_2$Flag_A_Infection_vs_Mock
)
)
attr(cstan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_1_3 <- dplyr::full_join(gstan_1_1_3[, c(1:4, 11:16)],
astan_1_1_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_1_3[is.nan(data.matrix(cstan_1_1_3))] <- NA
class(cstan_1_1_3) <- c("statRes", "data.frame")
attr(cstan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(cstan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(cstan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(sum(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 1] == 1))),
sum(length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 2] == 1))),
sum(length(which(flag_g_1_1_3[, 3] == 1)),
length(which(flag_a_1_1_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 1] == -1))),
sum(length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 2] == -1))),
sum(length(which(flag_g_1_1_3[, 3] == -1)),
length(which(flag_a_1_1_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_1_1_3, "statistical_test") <- "combined"
attr(cstan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_1_3, "pval_thresh") <- 0.05
attr(cstan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_1_3$e_data[, -1])) /
prod(dim(gfilta_1_1_3$e_data[, -1]))),
num_samps = dim(gfilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(cstan_1_1_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_1_1_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_1_3$e_data$Mass_Tag_ID,
mutant_vs_zombie = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
is.na(cstan_1_1_3$P_value_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
(cstan_1_1_3$P_value_A_mutant_vs_zombie > 0.05 &
cstan_1_1_3$P_value_G_mutant_vs_zombie < 0.05) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
!is.na(cstan_1_1_3$Flag_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_A_mutant_vs_zombie
),
mutant_vs_human = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_mutant_vs_human) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
is.na(cstan_1_1_3$P_value_A_mutant_vs_human) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
(cstan_1_1_3$P_value_A_mutant_vs_human > 0.05 &
cstan_1_1_3$P_value_G_mutant_vs_human < 0.05) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
!is.na(cstan_1_1_3$Flag_A_mutant_vs_human) ~
cstan_1_1_3$Flag_A_mutant_vs_human
),
zombie_vs_human = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_zombie_vs_human) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
is.na(cstan_1_1_3$P_value_A_zombie_vs_human) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
(cstan_1_1_3$P_value_A_zombie_vs_human > 0.05 &
cstan_1_1_3$P_value_G_zombie_vs_human < 0.05) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
!is.na(cstan_1_1_3$Flag_A_zombie_vs_human) ~
cstan_1_1_3$Flag_A_zombie_vs_human
)
)
attr(cstan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_1_3, "data_class") <- "pepData"
# main effects: 1; covariates: 2; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_2_3 <- dplyr::full_join(gstan_1_2_3[, c(1:4, 11:16)],
astan_1_2_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_2_3[is.nan(data.matrix(cstan_1_2_3))] <- NA
class(cstan_1_2_3) <- c("statRes", "data.frame")
attr(cstan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(cstan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(cstan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(sum(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 1] == 1))),
sum(length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 2] == 1))),
sum(length(which(flag_g_1_2_3[, 3] == 1)),
length(which(flag_a_1_2_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 1] == -1))),
sum(length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 2] == -1))),
sum(length(which(flag_g_1_2_3[, 3] == -1)),
length(which(flag_a_1_2_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_1_2_3, "statistical_test") <- "combined"
attr(cstan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_2_3, "pval_thresh") <- 0.05
attr(cstan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_2_3$e_data[, -1])) /
prod(dim(gfilta_1_2_3$e_data[, -1]))),
num_samps = dim(gfilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(cstan_1_2_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_1_2_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_2_3$e_data$Mass_Tag_ID,
mutant_vs_zombie = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
is.na(cstan_1_2_3$P_value_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
(cstan_1_2_3$P_value_A_mutant_vs_zombie > 0.05 &
cstan_1_2_3$P_value_G_mutant_vs_zombie < 0.05) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
!is.na(cstan_1_2_3$Flag_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_A_mutant_vs_zombie
),
mutant_vs_human = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_mutant_vs_human) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
is.na(cstan_1_2_3$P_value_A_mutant_vs_human) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
(cstan_1_2_3$P_value_A_mutant_vs_human > 0.05 &
cstan_1_2_3$P_value_G_mutant_vs_human < 0.05) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
!is.na(cstan_1_2_3$Flag_A_mutant_vs_human) ~
cstan_1_2_3$Flag_A_mutant_vs_human
),
zombie_vs_human = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_zombie_vs_human) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
is.na(cstan_1_2_3$P_value_A_zombie_vs_human) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
(cstan_1_2_3$P_value_A_zombie_vs_human > 0.05 &
cstan_1_2_3$P_value_G_zombie_vs_human < 0.05) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
!is.na(cstan_1_2_3$Flag_A_zombie_vs_human) ~
cstan_1_2_3$Flag_A_zombie_vs_human
)
)
attr(cstan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_2_3, "data_class") <- "pepData"
# main effects: 2; covariates: 0; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_0_3 <- dplyr::full_join(gstan_2_0_3[, c(1:4, 11:16)],
astan_2_0_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_0_3[is.nan(data.matrix(cstan_2_0_3))] <- NA
class(cstan_2_0_3) <- c("statRes", "data.frame")
attr(cstan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(cstan_2_0_3, "comparisons") <- c(
"Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"
)
attr(cstan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(sum(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 1] == 1))),
sum(length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 2] == 1))),
sum(length(which(flag_g_2_0_3[, 3] == 1)),
length(which(flag_a_2_0_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 1] == -1))),
sum(length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 2] == -1))),
sum(length(which(flag_g_2_0_3[, 3] == -1)),
length(which(flag_a_2_0_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_2_0_3, "statistical_test") <- "combined"
attr(cstan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_0_3, "pval_thresh") <- 0.05
attr(cstan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_0_3$e_data[, -1])) /
prod(dim(gfilta_2_0_3$e_data[, -1]))),
num_samps = dim(gfilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(cstan_2_0_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, NA, NA, 0, 0, 0, 0,
0, 0, NA, NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_0_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_0_3$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_0_3$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_0_3$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_0_3$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_none = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
is.na(cstan_2_0_3$P_value_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
(cstan_2_0_3$P_value_A_Infection_high_vs_Mock_none > 0.05 &
cstan_2_0_3$P_value_G_Infection_high_vs_Mock_none < 0.05) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
!is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none
),
Infection_low_vs_Mock_none = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
is.na(cstan_2_0_3$P_value_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
(cstan_2_0_3$P_value_A_Infection_low_vs_Mock_none > 0.05 &
cstan_2_0_3$P_value_G_Infection_low_vs_Mock_none < 0.05) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
!is.na(cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none
)
)
attr(cstan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_0_3, "data_class") <- "pepData"
# main effects: 2; covariates: 1; groups: 4 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_1_4 <- dplyr::full_join(gstan_2_1_4[, c(1:5, 16:27)],
astan_2_1_4) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_1_4[is.nan(data.matrix(cstan_2_1_4))] <- NA
class(cstan_2_1_4) <- c("statRes", "data.frame")
attr(cstan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(cstan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(cstan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(sum(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 1] == 1))),
sum(length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 2] == 1))),
sum(length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 3] == 1))),
sum(length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 4] == 1))),
sum(length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 5] == 1))),
sum(length(which(flag_g_2_1_4[, 6] == 1)),
length(which(flag_a_2_1_4[, 6] == 1)))),
Down_total = c(sum(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 1] == -1))),
sum(length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 2] == -1))),
sum(length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 3] == -1))),
sum(length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 4] == -1))),
sum(length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 5] == -1))),
sum(length(which(flag_g_2_1_4[, 6] == -1)),
length(which(flag_a_2_1_4[, 6] == -1)))),
Up_anova = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_gtest = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
row.names = NULL
)
attr(cstan_2_1_4, "statistical_test") <- "combined"
attr(cstan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_1_4, "pval_thresh") <- 0.05
attr(cstan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_1_4$e_data[, -1])) /
prod(dim(gfilta_2_1_4$e_data[, -1]))),
num_samps = dim(gfilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(cstan_2_1_4, "which_X") <-
c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
1, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0, 1,
NA, 0, 0, 0, 1, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_1_4, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_1_4$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_1_4$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_high > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Mock_high < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high
),
Infection_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low
),
Infection_low_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
is.na(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_high > 0.05 &
cstan_2_1_4$P_value_G_Infection_low_vs_Mock_high < 0.05) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
!is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high
),
Infection_low_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_low_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low
),
Mock_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
(cstan_2_1_4$P_value_A_Mock_high_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Mock_high_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low
)
)
attr(cstan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_1_4, "data_class") <- "pepData"
# main effects: 2; covariates: 2; groups: 4 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_2_4 <- dplyr::full_join(gstan_2_2_4[, c(1:5, 16:27)],
astan_2_2_4) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_2_4[is.nan(data.matrix(cstan_2_2_4))] <- NA
class(cstan_2_2_4) <- c("statRes", "data.frame")
attr(cstan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(cstan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(cstan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(sum(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 1] == 1))),
sum(length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 2] == 1))),
sum(length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 3] == 1))),
sum(length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 4] == 1))),
sum(length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 5] == 1))),
sum(length(which(flag_g_2_2_4[, 6] == 1)),
length(which(flag_a_2_2_4[, 6] == 1)))),
Down_total = c(sum(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 1] == -1))),
sum(length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 2] == -1))),
sum(length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 3] == -1))),
sum(length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 4] == -1))),
sum(length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 5] == -1))),
sum(length(which(flag_g_2_2_4[, 6] == -1)),
length(which(flag_a_2_2_4[, 6] == -1)))),
Up_anova = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_gtest = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
row.names = NULL
)
attr(cstan_2_2_4, "statistical_test") <- "combined"
attr(cstan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_2_4, "pval_thresh") <- 0.05
attr(cstan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_2_4$e_data[, -1])) /
prod(dim(gfilta_2_2_4$e_data[, -1]))),
num_samps = dim(gfilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(cstan_2_2_4, "which_X") <-
c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
1, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0, 1,
NA, 0, 0, 0, 1, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_2_4, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_2_4$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_2_4$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_high > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Mock_high < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high
),
Infection_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low
),
Infection_low_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
is.na(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_high > 0.05 &
cstan_2_2_4$P_value_G_Infection_low_vs_Mock_high < 0.05) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
!is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high
),
Infection_low_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_low_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low
),
Mock_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
(cstan_2_2_4$P_value_A_Mock_high_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Mock_high_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low
)
)
attr(cstan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_2_4, "data_class") <- "pepData"
# Save standards for IMD-ANOVA tests -------------------------------------------
# save() the following objects in "inst/testdata/standards_imd_anova.RData"
# astan_1_0_2,
# gstan_1_0_2,
# cstan_1_0_2,
# astan_1_1_3,
# gstan_1_1_3,
# cstan_1_1_3,
# astan_1_2_3,
# gstan_1_2_3,
# cstan_1_2_3,
# astan_2_0_3,
# gstan_2_0_3,
# cstan_2_0_3,
# astan_2_1_4,
# gstan_2_1_4,
# cstan_2_1_4,
# astan_2_2_4,
# gstan_2_2_4,
# cstan_2_2_4,
# tukey_pval_1_1_3,
# tukey_pval_1_2_3,
# tukey_pval_2_0_3,
# tukey_pval_2_1_4,
# tukey_pval_2_2_4,
# dunnett_1_1_3,
# dunnett_1_2_3,
# dunnett_2_0_3,
# dunnett_2_1_4,
# dunnett_2_2_4
# DO NOT PUT CODE HERE TO SAVE THE ABOVE OBJECTS. It may create problems with CRAN submission.
pmartR/pmartR documentation built on May 5, 2024, 12:03 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
# Load functions for calculating IMD-ANOVA standards ---------------------------
source(system.file("testdata",
"imd_anova_standard_fns.R",
package = "pmartR"))
# Load data and prepare omicsData objects --------------------------------------
load(system.file('testdata',
'little_pdata.RData',
package = 'pmartR'))
pdata <- as.pepData(e_data = edata,
f_data = fdata,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
# Create additional f_data objects with different main effects and covariates.
fdata2 <- fdata
set.seed(22250)
fdata2$Level <- c(sample(c("high", "low"), 9, replace = TRUE),
"none", "none", "none")
# Create a vector that has three different levels for Condition.
fdataZ <- fdata
set.seed(720)
fdataZ$Condition <- c(sample(c("zombie", "mutant"), 9, replace = TRUE),
"human", "human", "human")
fdataZ$Gender <- sample(c("F", "M"), 12, replace = TRUE)
fdataZ$Age <- round(runif(12, min = 19, max = 89), 2)
# Copy edata so the names of the samples can be changed.
edata2_2 <- edata
# Change some of the Infection samples to Mock samples.
names(edata2_2) <- c("Mass_Tag_ID",
paste0("Infection", 1:6),
paste0("Mock", 1:6))
# Create additional f_data objects with different main effects and covariates.
fdata2_2 <- fdata
# Update the sample names in f_data.
fdata2_2$SampleID <- c(paste0("Infection", 1:6),
paste0("Mock", 1:6))
# Update the first main effect to account for changing some infection samples to
# mock samples.
fdata2_2$Condition <- c(rep("Infection", 6),
rep("Mock", 6))
fdata2_2$Level <- c("high", "low", "high", "low", "high", "low", "high",
"high", "low", "low", "low", "high")
fdata2_2$Gender <- fdataZ$Gender
fdata2_2$Age <- fdataZ$Age
# Create pepData objects, log the data, add group_DF attributes ----------------
# NOTE: There are three numbers that follow each object's name. The first
# number represents the number of main effects. The second number represents
# the number of covariates. The third represents the number of groups (across
# all main effects).
# One main effect, zero covariates, two groups
pdata_1_0_2 <- pdata
pdata_1_0_2 <- edata_transform(pdata_1_0_2, "log")
pdata_1_0_2 <- group_designation(omicsData = pdata_1_0_2,
main_effects = "Condition")
groupDF_1_0_2 <- attr(pdata_1_0_2, "group_DF")
# One main effect, one covariate, three groups
pdata_1_1_3 <- as.pepData(e_data = edata,
f_data = fdataZ,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
pdata_1_1_3 <- edata_transform(pdata_1_1_3, "log")
pdata_1_1_3 <- group_designation(omicsData = pdata_1_1_3,
main_effects = "Condition",
covariates = "Gender")
groupDF_1_1_3 <- attr(pdata_1_1_3, "group_DF")
# One main effect, two covariates, three groups
pdata_1_2_3 <- as.pepData(e_data = edata,
f_data = fdataZ,
e_meta = emeta,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID",
emeta_cname = "Protein")
pdata_1_2_3 <- edata_transform(pdata_1_2_3, "log")
pdata_1_2_3 <- group_designation(omicsData = pdata_1_2_3,
main_effects = "Condition",
covariates = c("Gender", "Age"))
groupDF_1_2_3 <- attr(pdata_1_2_3, "group_DF")
# Two main effects, zero covariates, three groups
pdata_2_0_3 <- as.pepData(e_data = edata,
f_data = fdata2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_0_3 <- edata_transform(pdata_2_0_3, "log")
pdata_2_0_3 <- group_designation(omicsData = pdata_2_0_3,
main_effects = c("Condition", "Level"))
groupDF_2_0_3 <- attr(pdata_2_0_3, "group_DF")
# Two main effects, one covariate, four groups
pdata_2_1_4 <- as.pepData(e_data = edata2_2,
f_data = fdata2_2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_1_4 <- edata_transform(pdata_2_1_4, "log")
pdata_2_1_4 <- group_designation(omicsData = pdata_2_1_4,
main_effects = c("Condition", "Level"),
covariates = "Gender")
groupDF_2_1_4 <- attr(pdata_2_1_4, "group_DF")
# Two main effects, two covariates, four groups
pdata_2_2_4 <- as.pepData(e_data = edata2_2,
f_data = fdata2_2,
edata_cname = "Mass_Tag_ID",
fdata_cname = "SampleID")
pdata_2_2_4 <- edata_transform(pdata_2_2_4, "log")
pdata_2_2_4 <- group_designation(omicsData = pdata_2_2_4,
main_effects = c("Condition", "Level"),
covariates = c("Gender", "Age"))
groupDF_2_2_4 <- attr(pdata_2_2_4, "group_DF")
# Filter IMD-ANOVAly -----------------------------------------------------------
# One main effect, zero covariates, two groups
filta_1_0_2 <- imdanova_filter(pdata_1_0_2)
afilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_0_2 <- applyFilt(filta_1_0_2, pdata_1_0_2,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# One main effect, one covariate, three groups
filta_1_1_3 <- imdanova_filter(pdata_1_1_3)
afilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_1_3 <- applyFilt(filta_1_1_3, pdata_1_1_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# One main effect, two covariates, three groups
filta_1_2_3 <- imdanova_filter(pdata_1_2_3)
afilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_1_2_3 <- applyFilt(filta_1_2_3, pdata_1_2_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, zero covariates, three groups
filta_2_0_3 <- imdanova_filter(pdata_2_0_3)
afilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_0_3 <- applyFilt(filta_2_0_3, pdata_2_0_3,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, one covariate, four groups
filta_2_1_4 <- imdanova_filter(pdata_2_1_4)
afilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_1_4 <- applyFilt(filta_2_1_4, pdata_2_1_4,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Two main effects, two covariates, four groups
filta_2_2_4 <- imdanova_filter(pdata_2_2_4)
afilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_anova = 2,
remove_singleton_groups = FALSE)
gfilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
cfilta_2_2_4 <- applyFilt(filta_2_2_4, pdata_2_2_4,
min_nonmiss_anova = 2,
min_nonmiss_gtest = 2,
remove_singleton_groups = FALSE)
# Save filter objects ---------------------------------------------------------
# save(
#
# afilta_1_0_2,
# gfilta_1_0_2,
# cfilta_1_0_2,
#
# afilta_1_1_3,
# gfilta_1_1_3,
# cfilta_1_1_3,
#
# afilta_1_2_3,
# gfilta_1_2_3,
# cfilta_1_2_3,
#
# afilta_2_0_3,
# gfilta_2_0_3,
# cfilta_2_0_3,
#
# afilta_2_1_4,
# gfilta_2_1_4,
# cfilta_2_1_4,
#
# afilta_2_2_4,
# gfilta_2_2_4,
# cfilta_2_2_4,
#
# file = "/Users/mart077/pmartR/inst/testdata/standards_filter.RData"
# )
# Assemble ANOVA standards -----------------------------------------------------
Xmatrix_1_1_3 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1,
1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0), dim = c(12L, 4L), dimnames = list(
c("1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11",
"12"), c("(Intercept)", "Groupzombie", "Grouphuman", "GenderM"
)), assign = c(0L, 1L, 1L, 2L), contrasts = list(Group = "contr.treatment",
Gender = "contr.treatment"))
Xmatrix_1_2_3 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1,
1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57), dim = c(12L,
5L))
Xmatrix_2_1_4 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0), dim = c(12L, 4L))
# with interaction effect
Xmatrix_2_1_4_full <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,
0), dim = c(12L, 5L))
Xmatrix_2_2_4 <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57), dim = c(12L,
5L))
# with interaction effect
Xmatrix_2_2_4_full <- structure(c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0,
1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 59.68, 37.16, 69.4, 73.12, 69.27,
47.68, 53.79, 19.45, 27.06, 30.07, 22.04, 31.57, 0, 0, 0, 0,
0, 0, 0, 0, 1, 1, 1, 0), dim = c(12L, 6L))
# Selected Full/Reduced Model
which_X_2_1_4_a <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0))
which_X_2_2_4_a <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0))
which_X_2_1_4_g <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0))
which_X_2_2_4_g <- structure(c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0))
# main effects: 1; covariates: 0; groups: 2 ---------------
mean_a_1_0_2 <- data.frame(
Mean_Infection = rowMeans(afilta_1_0_2$e_data[, 2:10],
na.rm = TRUE),
Mean_Mock = rowMeans(afilta_1_0_2$e_data[, 11:13],
na.rm = TRUE)
)
pval_a_1_0_2 <- data.frame(
Infection_vs_Mock = a(data = afilta_1_0_2$e_data[, -1],
groups = attr(afilta_1_0_2, "group_DF")$Group)
)
flag_a_1_0_2 <- data.frame(
Infection_vs_Mock = aflag(
grp1 = mean_a_1_0_2$Mean_Infection,
grp2 = mean_a_1_0_2$Mean_Mock,
pvals = pval_a_1_0_2[, 1],
cutoff = 0.05
)
)
astan_1_0_2 <- data.frame(
Mass_Tag_ID = afilta_1_0_2$e_data$Mass_Tag_ID,
Count_Infection = unname(rowSums(!is.na(afilta_1_0_2$e_data[, 2:10]))),
Count_Mock = unname(rowSums(!is.na(afilta_1_0_2$e_data[, 11:13]))),
mean_a_1_0_2,
Fold_change_Infection_vs_Mock = (
mean_a_1_0_2[, 1] - mean_a_1_0_2[, 2]
),
P_value_A_Infection_vs_Mock = pval_a_1_0_2$Infection_vs_Mock,
Flag_A_Infection_vs_Mock = flag_a_1_0_2$Infection_vs_Mock,
row.names = NULL
)
class(astan_1_0_2) <- c("statRes", "data.frame")
attr(astan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(astan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(astan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_total = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_anova = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_anova = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_gtest = c(0),
Down_gtest = c(0),
row.names = NULL
)
attr(astan_1_0_2, "statistical_test") <- "anova"
attr(astan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(astan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(astan_1_0_2, "pval_thresh") <- 0.05
attr(astan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_0_2$e_data[, -1])) /
prod(dim(afilta_1_0_2$e_data[, -1]))),
num_samps = dim(afilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(astan_1_0_2, "which_X") <- rep(0, nrow(astan_1_0_2))
attr(astan_1_0_2, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_0_2$e_data$Mass_Tag_ID,
flag_a_1_0_2
)
attr(astan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
data_1_1_3 = afilta_1_1_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_1_3), Xmatrix = data.matrix(Xmatrix_1_1_3))
group_sampnames <- groupDF_1_1_3[,get_fdata_cname(afilta_1_1_3)]
groupData <- groupDF_1_1_3[group_sampnames %in% colnames(afilta_1_1_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_1_1_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", covariate_names = "Gender")
mean_a_1_1_3 <- get_lsmeans(data = data_1_1_3, xmatrix = Xmatrix_1_1_3, pred_grid = pred_grid, Betas = Betas)
beta_to_mu = pred_grid
beta_to_mu[,4] = 0
beta_to_mu = unique(beta_to_mu)
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
cmat = cmat %*% beta_to_mu
test_values <- get_test_values(afilta_1_1_3$e_data[, -1], Xmatrix_1_1_3, cmat)
group_counts_1_1_3 <- data.frame(
nona_mutant = rowSums(!is.na(data_1_1_3[, c(1, 3:4, 9)])),
nona_zombie = rowSums(!is.na(data_1_1_3[, c(2, 5:8)])),
nona_human = rowSums(!is.na(data_1_1_3[, 10:12]))
) %>%
dplyr::rowwise() %>%
dplyr::mutate(n_grp = sum(dplyr::c_across(nona_mutant:nona_human) == 0)) %>%
dplyr::ungroup()
nona_grps_1_1_3 <- rowSums(group_counts_1_1_3[, 1:3] != 0)
nona_counts_1_1_3 <- rowSums(!is.na(data_1_1_3))
# set mean_a_1_1_3 to NA for groups with zero observations
mean_a_1_1_3[group_counts_1_1_3[,-4] == 0] <- NA
diffs_1_1_3 <- mean_a_1_1_3 %>%
dplyr::mutate(
diff_m_z = Mean_mutant - Mean_zombie,
diff_m_h = Mean_mutant - Mean_human,
diff_z_h = Mean_zombie - Mean_human
) %>%
dplyr::select(diff_m_z, diff_m_h, diff_z_h)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3")
test_stat_1_1_3 <- diffs_1_1_3 %>%
cbind(group_counts_1_1_3) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_m_z = (diff_m_z / C1),
stat_m_h = (diff_m_h / C2),
stat_z_h = (diff_z_h / C3)
) %>%
dplyr::select(stat_m_z, stat_m_h, stat_z_h) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_1_1_3 <- test_stat_1_1_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = pt(
q = abs(stat_m_z),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_mutant_vs_human = pt(
q = abs(stat_m_h),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_zombie_vs_human = pt(
q = abs(stat_z_h),
df = nona_counts_1_1_3 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`row.names<-`(NULL)
flag_a_1_1_3 <- data.frame(
mutant_vs_zombie = aflag(
grp1 = mean_a_1_1_3$Mean_mutant,
grp2 = mean_a_1_1_3$Mean_zombie,
pvals = pval_a_1_1_3[, 1],
cutoff = 0.05
),
mutant_vs_human = aflag(
grp1 = mean_a_1_1_3$Mean_mutant,
grp2 = mean_a_1_1_3$Mean_human,
pvals = pval_a_1_1_3[, 2],
cutoff = 0.05
),
zombie_vs_human = aflag(
grp1 = mean_a_1_1_3$Mean_zombie,
grp2 = mean_a_1_1_3$Mean_human,
pvals = pval_a_1_1_3[, 3],
cutoff = 0.05
)
)
astan_1_1_3 <- data.frame(
Mass_Tag_ID = afilta_1_1_3$e_data$Mass_Tag_ID,
Count_mutant = group_counts_1_1_3$nona_mutant,
Count_zombie = group_counts_1_1_3$nona_zombie,
Count_human = group_counts_1_1_3$nona_human,
mean_a_1_1_3,
Fold_change_mutant_vs_zombie = (
mean_a_1_1_3[, 1] - mean_a_1_1_3[, 2]
),
Fold_change_mutant_vs_human = (
mean_a_1_1_3[, 1] - mean_a_1_1_3[, 3]
),
Fold_change_zombie_vs_human = (
mean_a_1_1_3[, 2] - mean_a_1_1_3[, 3]
),
pval_a_1_1_3,
Flag_A_mutant_vs_zombie = flag_a_1_1_3[, 1],
Flag_A_mutant_vs_human = flag_a_1_1_3[, 2],
Flag_A_zombie_vs_human = flag_a_1_1_3[, 3],
row.names = NULL
)
class(astan_1_1_3) <- c("statRes", "data.frame")
attr(astan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(astan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(astan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_total = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_1_1_3, "statistical_test") <- "anova"
attr(astan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(astan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(astan_1_1_3, "pval_thresh") <- 0.05
attr(astan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_1_3$e_data[, -1])) /
prod(dim(afilta_1_1_3$e_data[, -1]))),
num_samps = dim(afilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(astan_1_1_3, "which_X") <- rep(0, nrow(astan_1_1_3))
attr(astan_1_1_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_1_3$e_data$Mass_Tag_ID,
flag_a_1_1_3
)
attr(astan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_1_3, "data_class") <- "pepData"
tukey_stat_1_1_3 <- test_stat_1_1_3 * sqrt(2)
n_tests_1_1_3 <- rowSums(!is.na(tukey_stat_1_1_3))
suppressWarnings(
tukey_pval_1_1_3 <- tukey_stat_1_1_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = ptukey(
abs(stat_m_z),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
),
P_value_A_mutant_vs_human = ptukey(
abs(stat_m_h),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
),
P_value_A_zombie_vs_human = ptukey(
abs(stat_z_h),
nranges = 1,
nmeans = n_tests_1_1_3,
df = nona_counts_1_1_3 - n_tests_1_1_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_1_1_3 <- which(rowSums(!is.na(tukey_stat_1_1_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_1_1_3 <- which(!is.na(pval_a_1_1_3[just_one_1_1_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_1_1_3[just_one_1_1_3, ][not_na_1_1_3] <- data.matrix(
pval_a_1_1_3[just_one_1_1_3, ]
)[not_na_1_1_3]
# Convert back to a data frame.
tukey_pval_1_1_3 <- data.frame(tukey_pval_1_1_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
dunnett_1_1_3 <- pval_a_1_1_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_1_1_3 - no_nas) %>%
cbind(test_stat_1_1_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_m_z = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_z),
no_nas),
rep(abs(stat_m_z),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_m_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_h),
no_nas),
rep(abs(stat_m_h),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_z_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_z_h),
no_nas),
rep(abs(stat_z_h),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_m_z, pval_m_h, pval_z_h)
# main effects: 1; covariates: 2; groups: 3 ---------------
data_1_2_3 = afilta_1_2_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_2_3), Xmatrix = data.matrix(Xmatrix_1_2_3))
group_sampnames <- groupDF_1_2_3[,get_fdata_cname(afilta_1_2_3)]
groupData <- groupDF_1_2_3[group_sampnames %in% colnames(afilta_1_2_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_1_2_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", c("Gender", "Age"))
mean_a_1_2_3 <- get_lsmeans(data = data_1_2_3, xmatrix = Xmatrix_1_2_3, pred_grid = pred_grid, Betas = Betas, continuous_covar_inds = 5)
beta_to_mu = pred_grid
beta_to_mu[,4] = 0
beta_to_mu = unique(beta_to_mu)
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
cmat = cmat %*% beta_to_mu
group_counts_1_2_3 <- data.frame(
nona_mutant = rowSums(!is.na(data_1_2_3[, c(1, 3:4, 9)])),
nona_zombie = rowSums(!is.na(data_1_2_3[, c(2, 5:8)])),
nona_human = rowSums(!is.na(data_1_2_3[, 10:12]))
) %>%
dplyr::rowwise() %>%
dplyr::mutate(n_grp = sum(dplyr::c_across(nona_mutant:nona_human) == 0)) %>%
dplyr::ungroup()
nona_grps_1_2_3 <- rowSums(group_counts_1_2_3[, 1:3] != 0)
nona_counts_1_2_3 <- rowSums(!is.na(data_1_2_3))
mean_a_1_2_3[group_counts_1_2_3[,-4] == 0] <- NA
diffs_1_2_3 <- mean_a_1_2_3 %>%
dplyr::mutate(
diff_m_z = Mean_mutant - Mean_zombie,
diff_m_h = Mean_mutant - Mean_human,
diff_z_h = Mean_zombie - Mean_human
) %>%
dplyr::select(diff_m_z, diff_m_h, diff_z_h)
test_values <- get_test_values(afilta_1_2_3$e_data[, -1], Xmatrix_1_2_3, cmat)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3")
suppressWarnings(
test_stat_1_2_3 <- diffs_1_2_3 %>%
cbind(diff_denoms) %>%
cbind(group_counts_1_2_3) %>%
dplyr::mutate(
stat_m_z = (diff_m_z / C1),
stat_m_h = (diff_m_h / C2),
stat_z_h = (diff_z_h / C3)
) %>%
dplyr::select(stat_m_z, stat_m_h, stat_z_h) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
)
pval_a_1_2_3 <- test_stat_1_2_3 %>%
dplyr::mutate(
lg = group_counts_1_2_3$n_grp,
P_value_A_mutant_vs_zombie = pt(
q = abs(stat_m_z),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_mutant_vs_human = pt(
q = abs(stat_m_h),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_zombie_vs_human = pt(
q = abs(stat_z_h),
df = nona_counts_1_2_3 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`row.names<-`(NULL)
flag_a_1_2_3 <- data.frame(
mutant_vs_zombie = aflag(
grp1 = mean_a_1_2_3$Mean_mutant,
grp2 = mean_a_1_2_3$Mean_zombie,
pvals = pval_a_1_2_3[, 1],
cutoff = 0.05
),
mutant_vs_human = aflag(
grp1 = mean_a_1_2_3$Mean_mutant,
grp2 = mean_a_1_2_3$Mean_human,
pvals = pval_a_1_2_3[, 2],
cutoff = 0.05
),
zombie_vs_human = aflag(
grp1 = mean_a_1_2_3$Mean_zombie,
grp2 = mean_a_1_2_3$Mean_human,
pvals = pval_a_1_2_3[, 3],
cutoff = 0.05
)
)
astan_1_2_3 <- data.frame(
Mass_Tag_ID = afilta_1_2_3$e_data$Mass_Tag_ID,
Count_mutant = group_counts_1_2_3$nona_mutant,
Count_zombie = group_counts_1_2_3$nona_zombie,
Count_human = group_counts_1_2_3$nona_human,
mean_a_1_2_3,
Fold_change_mutant_vs_zombie = (
mean_a_1_2_3[, 1] - mean_a_1_2_3[, 2]
),
Fold_change_mutant_vs_human = (
mean_a_1_2_3[, 1] - mean_a_1_2_3[, 3]
),
Fold_change_zombie_vs_human = (
mean_a_1_2_3[, 2] - mean_a_1_2_3[, 3]
),
pval_a_1_2_3,
Flag_A_mutant_vs_zombie = flag_a_1_2_3[, 1],
Flag_A_mutant_vs_human = flag_a_1_2_3[, 2],
Flag_A_zombie_vs_human = flag_a_1_2_3[, 3],
row.names = NULL
)
class(astan_1_2_3) <- c("statRes", "data.frame")
attr(astan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(astan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(astan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_total = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_1_2_3, "statistical_test") <- "anova"
attr(astan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(astan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(astan_1_2_3, "pval_thresh") <- 0.05
attr(astan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_1_2_3$e_data[, -1])) /
prod(dim(afilta_1_2_3$e_data[, -1]))),
num_samps = dim(afilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(astan_1_2_3, "which_X") <- rep(0, nrow(astan_1_2_3))
attr(astan_1_2_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_1_2_3$e_data$Mass_Tag_ID,
flag_a_1_2_3
)
attr(astan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_1_2_3, "data_class") <- "pepData"
tukey_stat_1_2_3 <- test_stat_1_2_3 * sqrt(2)
n_tests_1_2_3 <- rowSums(!is.na(tukey_stat_1_2_3))
suppressWarnings(
tukey_pval_1_2_3 <- tukey_stat_1_2_3 %>%
dplyr::mutate(
P_value_A_mutant_vs_zombie = ptukey(
abs(stat_m_z),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
),
P_value_A_mutant_vs_human = ptukey(
abs(stat_m_h),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
),
P_value_A_zombie_vs_human = ptukey(
abs(stat_z_h),
nranges = 1,
nmeans = n_tests_1_2_3,
df = nona_counts_1_2_3 - n_tests_1_2_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_mutant_vs_zombie,
P_value_A_mutant_vs_human,
P_value_A_zombie_vs_human) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_1_2_3 <- which(rowSums(!is.na(tukey_stat_1_2_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_1_2_3 <- which(!is.na(pval_a_1_2_3[just_one_1_2_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_1_2_3[just_one_1_2_3, ][not_na_1_2_3] <- data.matrix(
pval_a_1_2_3[just_one_1_2_3, ]
)[not_na_1_2_3]
# Convert back to a data frame.
tukey_pval_1_2_3 <- data.frame(tukey_pval_1_2_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
suppressWarnings(
dunnett_1_2_3 <- pval_a_1_2_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_1_2_3 - no_nas) %>%
cbind(test_stat_1_2_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_m_z = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_z),
no_nas),
rep(abs(stat_m_z),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_m_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_m_h),
no_nas),
rep(abs(stat_m_h),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_z_h = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_z_h),
no_nas),
rep(abs(stat_z_h),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_m_z, pval_m_h, pval_z_h)
)
# main effects: 2; covariates: 0; groups: 3 ---------------
# no covariates? should just be the observed means.
mean_a_2_0_3 <- data.frame(
Mean_Infection_high = rowMeans(afilta_2_0_3$e_data[, c(2:4, 7:8)],
na.rm = TRUE),
Mean_Infection_low = rowMeans(afilta_2_0_3$e_data[, c(5:6, 9:10)],
na.rm = TRUE),
Mean_Mock_none = rowMeans(afilta_2_0_3$e_data[, 11:13],
na.rm = TRUE)
)
group_counts_2_0_3 <- data.frame(
nona_iHigh = rowSums(!is.na(afilta_2_0_3$e_data[, c(2:4, 7:8)])),
nona_iLow = rowSums(!is.na(afilta_2_0_3$e_data[, c(5:6, 9:10)])),
nona_mNone = rowSums(!is.na(afilta_2_0_3$e_data[, 11:13]))
)
nona_grps_2_0_3 <- rowSums(group_counts_2_0_3 != 0)
sigma_2_0_3 <- dplyr::rowwise(afilta_2_0_3$e_data[, -1]) %>%
dplyr::mutate(
stdev = summary(lm(
dplyr::c_across(Infection1:Mock3) ~ groupDF_2_0_3$Group
))$sigma
) %>%
dplyr::pull(stdev)
nona_counts_2_0_3 <- rowSums(!is.na(afilta_2_0_3$e_data[, -1]))
diffs_2_0_3 <- mean_a_2_0_3 %>%
dplyr::mutate(
diff_high_low = Mean_Infection_high - Mean_Infection_low,
diff_high_none = Mean_Infection_high - Mean_Mock_none,
diff_low_none = Mean_Infection_low - Mean_Mock_none
) %>%
dplyr::select(diff_high_low, diff_high_none, diff_low_none)
test_stat_2_0_3 <- diffs_2_0_3 %>%
dplyr::mutate(
stat_high_low = (diff_high_low /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 1] +
1/group_counts_2_0_3[, 2]))),
stat_high_none = (diff_high_none /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 1] +
1/group_counts_2_0_3[, 3]))),
stat_low_none = (diff_low_none /
(sigma_2_0_3 * sqrt(1/group_counts_2_0_3[, 2] +
1/group_counts_2_0_3[, 3])))
) %>%
dplyr::select(stat_high_low, stat_high_none, stat_low_none)
pval_a_2_0_3 <- test_stat_2_0_3 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_high_low),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_none = pt(
q = abs(stat_high_none),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_none = pt(
q = abs(stat_low_none),
df = nona_counts_2_0_3 - nona_grps_2_0_3,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_none,
P_value_A_Infection_low_vs_Mock_none)
flag_a_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_high,
grp2 = mean_a_2_0_3$Mean_Infection_low,
pvals = pval_a_2_0_3[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_none = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_high,
grp2 = mean_a_2_0_3$Mean_Mock_none,
pvals = pval_a_2_0_3[, 2],
cutoff = 0.05
),
Infection_low_vs_Mock_none = aflag(
grp1 = mean_a_2_0_3$Mean_Infection_low,
grp2 = mean_a_2_0_3$Mean_Mock_none,
pvals = pval_a_2_0_3[, 3],
cutoff = 0.05
)
)
astan_2_0_3 <- data.frame(
Mass_Tag_ID = afilta_2_0_3$e_data$Mass_Tag_ID,
Count_Infection_high = unname(
rowSums(!is.na(afilta_2_0_3$e_data[, c(2:4, 7:8)]))
),
Count_Infection_low = unname(
rowSums(!is.na(afilta_2_0_3$e_data[, c(5:6, 9:10)]))
),
Count_Mock_none = unname(rowSums(!is.na(afilta_2_0_3$e_data[, 11:13]))),
mean_a_2_0_3,
Fold_change_Infection_high_vs_Infection_low = (
mean_a_2_0_3[, 1] - mean_a_2_0_3[, 2]
),
Fold_change_Infection_high_vs_Mock_none = (
mean_a_2_0_3[, 1] - mean_a_2_0_3[, 3]
),
Fold_change_Infection_low_vs_Mock_none = (
mean_a_2_0_3[, 2] - mean_a_2_0_3[, 3]
),
pval_a_2_0_3,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_0_3[, 1],
Flag_A_Infection_high_vs_Mock_none = flag_a_2_0_3[, 2],
Flag_A_Infection_low_vs_Mock_none = flag_a_2_0_3[, 3],
row.names = NULL
)
class(astan_2_0_3) <- c("statRes", "data.frame")
attr(astan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(astan_2_0_3, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none")
attr(astan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_total = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_anova = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_gtest = c(0, 0, 0),
Down_gtest = c(0, 0, 0),
row.names = NULL
)
attr(astan_2_0_3, "statistical_test") <- "anova"
attr(astan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(astan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(astan_2_0_3, "pval_thresh") <- 0.05
attr(astan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_0_3$e_data[, -1])) /
prod(dim(afilta_2_0_3$e_data[, -1]))),
num_samps = dim(afilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(astan_2_0_3, "which_X") <- rep(0, nrow(astan_2_0_3))
attr(astan_2_0_3, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_0_3$e_data$Mass_Tag_ID,
flag_a_2_0_3
)
attr(astan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_0_3, "data_class") <- "pepData"
tukey_stat_2_0_3 <- test_stat_2_0_3 * sqrt(2)
n_tests_2_0_3 <- rowSums(!is.na(tukey_stat_2_0_3))
suppressWarnings(
tukey_pval_2_0_3 <- tukey_stat_2_0_3 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_high_low),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_none = ptukey(
abs(stat_high_none),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_none = ptukey(
abs(stat_low_none),
nranges = 1,
nmeans = n_tests_2_0_3,
df = nona_counts_2_0_3 - n_tests_2_0_3,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_none,
P_value_A_Infection_low_vs_Mock_none) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_0_3 <- which(rowSums(!is.na(tukey_stat_2_0_3)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_0_3 <- which(!is.na(pval_a_2_0_3[just_one_2_0_3, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_0_3[just_one_2_0_3, ][not_na_2_0_3] <- data.matrix(
pval_a_2_0_3[just_one_2_0_3, ]
)[not_na_2_0_3]
# Convert back to a data frame.
tukey_pval_2_0_3 <- data.frame(tukey_pval_2_0_3)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(5)
dunnett_2_0_3 <- pval_a_2_0_3 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_0_3 - no_nas) %>%
cbind(test_stat_2_0_3) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_high_low = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_high_low),
no_nas),
rep(abs(stat_high_low),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_high_none = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_high_none),
no_nas),
rep(abs(stat_high_none),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_low_none = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_low_none),
no_nas),
rep(abs(stat_low_none),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_high_low, pval_high_none, pval_low_none)
# main effects: 2; covariates: 1; groups: 4 ---------------
data_2_1_4 = data.matrix(afilta_2_1_4$e_data[, -1])
covariate_names = colnames(attr(attr(afilta_2_1_4, "group_DF"), "covariates"))[-1]
main_effect_names = attr(attr(afilta_2_1_4, "group_DF"), "main_effects")
group_sampnames <- groupDF_2_1_4[,get_fdata_cname(afilta_2_1_4)]
groupData <- groupDF_2_1_4[group_sampnames %in% colnames(afilta_2_1_4$e_data),]
groupData[,main_effect_names] <- lapply(groupData[main_effect_names], function(x) factor(x, levels=unique(x)))
groupData <- groupData %>%
dplyr::left_join(afilta_2_1_4$f_data)
# no continuous covariates
pred_grid_red_2_1_4 = get_pred_grid(groupData, c("Condition", "Level"), "Gender")
pred_grid_full_2_1_4 = get_pred_grid(groupData, c("Condition", "Level"), "Gender", as.formula("~Condition*Level+Gender"))
cobra <- run_twofactor_cpp(
data = data.matrix(data_2_1_4),
gpData = groupData[,c("Group", main_effect_names, covariate_names)],
Xfull=Xmatrix_2_1_4_full, Xred = Xmatrix_2_1_4,
pred_grid_full = pred_grid_full_2_1_4, pred_grid_red = pred_grid_red_2_1_4,
continuous_covar_inds = numeric(0)
)
group_counts_2_1_4 <- data.frame(
nona_Infection_high = rowSums(!is.na(data_2_1_4[, c(1, 3, 5)])),
nona_Infection_low = rowSums(!is.na(data_2_1_4[, c(2, 4, 6)])),
nona_Mock_high = rowSums(!is.na(data_2_1_4[, c(7, 8, 12)])),
nona_Mock_low = rowSums(!is.na(data_2_1_4[, c(9, 10, 11)]))
) %>%
dplyr::ungroup()
nona_grps_2_1_4 <- unname(rowSums(group_counts_2_1_4 != 0))
nona_counts_2_1_4 <- unname(rowSums(!is.na(data_2_1_4)))
mean_a_2_1_4 <- data.frame(cobra$lsmeans)
colnames(mean_a_2_1_4) <- paste0("Mean_", unique(groupData$Group))
diffs_2_1_4 <- mean_a_2_1_4 %>%
dplyr::mutate(
diff_ih_il = Mean_Infection_high - Mean_Infection_low,
diff_ih_mh = Mean_Infection_high - Mean_Mock_high,
diff_ih_ml = Mean_Infection_high - Mean_Mock_low,
diff_il_mh = Mean_Infection_low - Mean_Mock_high,
diff_il_ml = Mean_Infection_low - Mean_Mock_low,
diff_mh_ml = Mean_Mock_high - Mean_Mock_low
) %>%
dplyr::select(diff_ih_il, diff_ih_mh, diff_ih_ml,
diff_il_mh, diff_il_ml, diff_mh_ml)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
beta_to_mu <- pred_grid_red_2_1_4
beta_to_mu[,4] <- 0
beta_to_mu <- unique(beta_to_mu)
cmat_red <- cmat %*% beta_to_mu
beta_to_mu_full <- pred_grid_full_2_1_4
beta_to_mu_full[,4] <- 0
beta_to_mu_full <- unique(beta_to_mu_full)
cmat_full <- cmat %*% beta_to_mu_full
test_values <- get_test_values_twofactor(data_2_1_4, Xmatrix_2_1_4, Xmatrix_2_1_4_full, cmat_red, cmat_full, cobra$which_X)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3", "C4", "C5", "C6")
test_stat_2_1_4 <- diffs_2_1_4 %>%
cbind(group_counts_2_1_4) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_ih_il = (diff_ih_il / C1),
stat_ih_mh = (diff_ih_mh / C2),
stat_ih_ml = (diff_ih_ml / C3),
stat_il_mh = (diff_il_mh / C4),
stat_il_ml = (diff_il_ml / C5),
stat_mh_ml = (diff_mh_ml / C6)
) %>%
dplyr::select(stat_ih_il, stat_ih_mh, stat_ih_ml,
stat_il_mh, stat_il_ml, stat_mh_ml) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_2_1_4 <- test_stat_2_1_4 %>%
dplyr::mutate(
lg = group_counts_2_1_4$n_grp,
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_ih_il),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_high = pt(
q = abs(stat_ih_mh),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_low = pt(
q = abs(stat_ih_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_high = pt(
q = abs(stat_il_mh),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_low = pt(
q = abs(stat_il_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Mock_high_vs_Mock_low = pt(
q = abs(stat_mh_ml),
df = nona_counts_2_1_4 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low) %>%
`row.names<-`(NULL)
flag_a_2_1_4 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,2],
pvals = pval_a_2_1_4[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_high = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,3],
pvals = pval_a_2_1_4[, 2],
cutoff = 0.05
),
Infection_high_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,1],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 3],
cutoff = 0.05
),
Infection_low_vs_Mock_high = aflag(
grp1 = mean_a_2_1_4[,2],
grp2 = mean_a_2_1_4[,3],
pvals = pval_a_2_1_4[, 4],
cutoff = 0.05
),
Infection_low_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,2],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 5],
cutoff = 0.05
),
Mock_high_vs_Mock_low = aflag(
grp1 = mean_a_2_1_4[,3],
grp2 = mean_a_2_1_4[,4],
pvals = pval_a_2_1_4[, 6],
cutoff = 0.05
)
)
astan_2_1_4 <- data.frame(
Mass_Tag_ID = afilta_2_1_4$e_data$Mass_Tag_ID,
Count_Infection_high = group_counts_2_1_4$nona_Infection_high,
Count_Infection_low = group_counts_2_1_4$nona_Infection_low,
Count_Mock_high = group_counts_2_1_4$nona_Mock_high,
Count_Mock_low = group_counts_2_1_4$nona_Mock_low,
mean_a_2_1_4,
Fold_change_Infection_high_vs_Infection_low = diffs_2_1_4$diff_ih_il,
Fold_change_Infection_high_vs_Mock_high = diffs_2_1_4$diff_ih_mh,
Fold_change_Infection_high_vs_Mock_low = diffs_2_1_4$diff_ih_ml,
Fold_change_Infection_low_vs_Mock_high = diffs_2_1_4$diff_il_mh,
Fold_change_Infection_low_vs_Mock_low = diffs_2_1_4$diff_il_ml,
Fold_change_Mock_high_vs_Mock_low = diffs_2_1_4$diff_mh_ml,
pval_a_2_1_4,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_1_4[, 1],
Flag_A_Infection_high_vs_Mock_high = flag_a_2_1_4[, 2],
Flag_A_Infection_high_vs_Mock_low = flag_a_2_1_4[, 3],
Flag_A_Infection_low_vs_Mock_high = flag_a_2_1_4[, 4],
Flag_A_Infection_low_vs_Mock_low = flag_a_2_1_4[, 5],
Flag_A_Mock_high_vs_Mock_low = flag_a_2_1_4[, 6],
row.names = NULL
)
class(astan_2_1_4) <- c("statRes", "data.frame")
attr(astan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(astan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(astan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_total = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_anova = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_gtest = c(0, 0, 0, 0, 0, 0),
Down_gtest = c(0, 0, 0, 0, 0, 0),
row.names = NULL
)
attr(astan_2_1_4, "statistical_test") <- "anova"
attr(astan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(astan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(astan_2_1_4, "pval_thresh") <- 0.05
attr(astan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_1_4$e_data[, -1])) /
prod(dim(afilta_2_1_4$e_data[, -1]))),
num_samps = dim(afilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(astan_2_1_4, "which_X") <- which_X_2_1_4_a
attr(astan_2_1_4, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_1_4$e_data$Mass_Tag_ID,
flag_a_2_1_4
)
attr(astan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_1_4, "data_class") <- "pepData"
tukey_stat_2_1_4 <- test_stat_2_1_4 * sqrt(2)
n_tests_2_1_4 <- rowSums(!is.na(tukey_stat_2_1_4))
suppressWarnings(
tukey_pval_2_1_4 <- tukey_stat_2_1_4 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_ih_il),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_high = ptukey(
abs(stat_ih_mh),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_low = ptukey(
abs(stat_ih_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_high = ptukey(
abs(stat_il_mh),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_low = ptukey(
abs(stat_il_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
),
P_value_A_Mock_high_vs_Mock_low = ptukey(
abs(stat_mh_ml),
nranges = 1,
nmeans = n_tests_2_1_4,
df = nona_counts_2_1_4 - n_tests_2_1_4,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low,) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_1_4 <- which(rowSums(!is.na(tukey_stat_2_1_4)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_1_4 <- which(!is.na(pval_a_2_1_4[just_one_2_1_4, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_1_4[just_one_2_1_4, ][not_na_2_1_4] <- data.matrix(
pval_a_2_1_4[just_one_2_1_4, ]
)[not_na_2_1_4]
# Convert back to a data frame.
tukey_pval_2_1_4 <- data.frame(tukey_pval_2_1_4)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
dunnett_2_1_4 <- pval_a_2_1_4 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_1_4 - no_nas) %>%
cbind(test_stat_2_1_4) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_ih_il = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_il),
no_nas),
rep(abs(stat_ih_il),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_mh),
no_nas),
rep(abs(stat_ih_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_ml),
no_nas),
rep(abs(stat_ih_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_mh),
no_nas),
rep(abs(stat_il_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_ml),
no_nas),
rep(abs(stat_il_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_mh_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_mh_ml),
no_nas),
rep(abs(stat_mh_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_ih_il, pval_ih_mh, pval_ih_ml,
pval_il_mh, pval_il_ml, pval_mh_ml)
# main effects: 2; covariates: 2; groups: 4 ---------------
data_2_2_4 = data.matrix(afilta_2_2_4$e_data[, -1])
covariate_names = colnames(attr(attr(afilta_2_2_4, "group_DF"), "covariates"))[-1]
main_effect_names = attr(attr(afilta_2_2_4, "group_DF"), "main_effects")
group_sampnames <- groupDF_2_2_4[,get_fdata_cname(afilta_2_2_4)]
groupData <- groupDF_2_2_4[group_sampnames %in% colnames(afilta_2_2_4$e_data),]
groupData <- groupData %>%
dplyr::left_join(afilta_2_2_4$f_data)
groupData[,main_effect_names] <- lapply(groupData[main_effect_names], function(x) factor(x, levels=unique(x)))
pred_grid_red_2_2_4 = get_pred_grid(groupData, main_effect_names, covariate_names)
pred_grid_full_2_2_4 = get_pred_grid(groupData, main_effect_names, covariate_names, fspec = as.formula("~Condition*Level+Gender+Age"))
cobra <- run_twofactor_cpp(
data = data.matrix(data_2_2_4),
gpData = groupData[,c("Group", main_effect_names, covariate_names)],
Xfull=Xmatrix_2_2_4_full, Xred = Xmatrix_2_2_4,
pred_grid_full = pred_grid_full_2_2_4, pred_grid_red = pred_grid_red_2_2_4,
continuous_covar_inds = 5
)
group_counts_2_2_4 <- data.frame(
nona_Infection_high = rowSums(!is.na(data_2_2_4[, c(1, 3, 5)])),
nona_Infection_low = rowSums(!is.na(data_2_2_4[, c(2, 4, 6)])),
nona_Mock_high = rowSums(!is.na(data_2_2_4[, c(7, 8, 12)])),
nona_Mock_low = rowSums(!is.na(data_2_2_4[, c(9, 10, 11)]))
) %>%
dplyr::ungroup()
nona_grps_2_2_4 <- unname(rowSums(group_counts_2_2_4 != 0))
nona_counts_2_2_4 <- unname(rowSums(!is.na(data_2_2_4)))
mean_a_2_2_4 <- data.frame(cobra$lsmeans)
colnames(mean_a_2_2_4) <- paste0("Mean_", unique(groupData$Group))
diffs_2_2_4 <- mean_a_2_2_4 %>%
dplyr::mutate(
diff_ih_il = Mean_Infection_high - Mean_Infection_low,
diff_ih_mh = Mean_Infection_high - Mean_Mock_high,
diff_ih_ml = Mean_Infection_high - Mean_Mock_low,
diff_il_mh = Mean_Infection_low - Mean_Mock_high,
diff_il_ml = Mean_Infection_low - Mean_Mock_low,
diff_mh_ml = Mean_Mock_high - Mean_Mock_low
) %>%
dplyr::select(diff_ih_il, diff_ih_mh, diff_ih_ml,
diff_il_mh, diff_il_ml, diff_mh_ml)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
beta_to_mu = pred_grid_red_2_2_4
beta_to_mu[,4:5] <- 0
beta_to_mu <- unique(beta_to_mu)
cmat_red <- cmat %*% beta_to_mu
beta_to_mu_full = pred_grid_full_2_2_4
beta_to_mu_full[,4:5] <- 0
beta_to_mu_full <- unique(beta_to_mu_full)
cmat_full <- cmat %*% beta_to_mu_full
test_values <- get_test_values_twofactor(data_2_2_4, Xmatrix_2_2_4, Xmatrix_2_2_4_full, cmat_red, cmat_full, cobra$which_X)
diff_denoms <- test_values$diff_denoms
colnames(diff_denoms) <- c("C1", "C2", "C3", "C4", "C5", "C6")
test_stat_2_2_4 <- diffs_2_2_4 %>%
cbind(group_counts_2_2_4) %>%
cbind(diff_denoms) %>%
dplyr::mutate(
stat_ih_il = (diff_ih_il / C1),
stat_ih_mh = (diff_ih_mh / C2),
stat_ih_ml = (diff_ih_ml / C3),
stat_il_mh = (diff_il_mh / C4),
stat_il_ml = (diff_il_ml / C5),
stat_mh_ml = (diff_mh_ml / C6)
) %>%
dplyr::select(stat_ih_il, stat_ih_mh, stat_ih_ml,
stat_il_mh, stat_il_ml, stat_mh_ml) %>%
dplyr::ungroup() %>%
`row.names<-`(NULL)
pval_a_2_2_4 <- test_stat_2_2_4 %>%
dplyr::mutate(
lg = group_counts_2_2_4$n_grp,
P_value_A_Infection_high_vs_Infection_low = pt(
q = abs(stat_ih_il),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_high = pt(
q = abs(stat_ih_mh),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_high_vs_Mock_low = pt(
q = abs(stat_ih_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_high = pt(
q = abs(stat_il_mh),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Infection_low_vs_Mock_low = pt(
q = abs(stat_il_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2,
P_value_A_Mock_high_vs_Mock_low = pt(
q = abs(stat_mh_ml),
df = nona_counts_2_2_4 - test_values$ranks,
lower.tail = FALSE
) * 2
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low) %>%
`row.names<-`(NULL)
flag_a_2_2_4 <- data.frame(
Infection_high_vs_Infection_low = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,2],
pvals = pval_a_2_2_4[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_high = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,3],
pvals = pval_a_2_2_4[, 2],
cutoff = 0.05
),
Infection_high_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,1],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 3],
cutoff = 0.05
),
Infection_low_vs_Mock_high = aflag(
grp1 = mean_a_2_2_4[,2],
grp2 = mean_a_2_2_4[,3],
pvals = pval_a_2_2_4[, 4],
cutoff = 0.05
),
Infection_low_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,2],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 5],
cutoff = 0.05
),
Mock_high_vs_Mock_low = aflag(
grp1 = mean_a_2_2_4[,3],
grp2 = mean_a_2_2_4[,4],
pvals = pval_a_2_2_4[, 6],
cutoff = 0.05
)
)
astan_2_2_4 <- data.frame(
Mass_Tag_ID = afilta_2_2_4$e_data$Mass_Tag_ID,
Count_Infection_high = group_counts_2_2_4$nona_Infection_high,
Count_Infection_low = group_counts_2_2_4$nona_Infection_low,
Count_Mock_high = group_counts_2_2_4$nona_Mock_high,
Count_Mock_low = group_counts_2_2_4$nona_Mock_low,
mean_a_2_2_4,
Fold_change_Infection_high_vs_Infection_low = diffs_2_2_4$diff_ih_il,
Fold_change_Infection_high_vs_Mock_high = diffs_2_2_4$diff_ih_mh,
Fold_change_Infection_high_vs_Mock_low = diffs_2_2_4$diff_ih_ml,
Fold_change_Infection_low_vs_Mock_high = diffs_2_2_4$diff_il_mh,
Fold_change_Infection_low_vs_Mock_low = diffs_2_2_4$diff_il_ml,
Fold_change_Mock_high_vs_Mock_low = diffs_2_2_4$diff_mh_ml,
pval_a_2_2_4,
Flag_A_Infection_high_vs_Infection_low = flag_a_2_2_4[, 1],
Flag_A_Infection_high_vs_Mock_high = flag_a_2_2_4[, 2],
Flag_A_Infection_high_vs_Mock_low = flag_a_2_2_4[, 3],
Flag_A_Infection_low_vs_Mock_high = flag_a_2_2_4[, 4],
Flag_A_Infection_low_vs_Mock_low = flag_a_2_2_4[, 5],
Flag_A_Mock_high_vs_Mock_low = flag_a_2_2_4[, 6],
row.names = NULL
)
class(astan_2_2_4) <- c("statRes", "data.frame")
attr(astan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(astan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(astan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_total = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_anova = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_gtest = c(0, 0, 0, 0, 0, 0),
Down_gtest = c(0, 0, 0, 0, 0, 0),
row.names = NULL
)
attr(astan_2_2_4, "statistical_test") <- "anova"
attr(astan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(astan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(astan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(astan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(astan_2_2_4, "pval_thresh") <- 0.05
attr(astan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(afilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(afilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(afilta_2_2_4$e_data[, -1])) /
prod(dim(afilta_2_2_4$e_data[, -1]))),
num_samps = dim(afilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(astan_2_2_4, "which_X") <- which_X_2_2_4_a
attr(astan_2_2_4, "bpFlags") <- data.frame(
Mass_Tag_ID = afilta_2_2_4$e_data$Mass_Tag_ID,
flag_a_2_2_4
)
attr(astan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(astan_2_2_4, "data_class") <- "pepData"
tukey_stat_2_2_4 <- test_stat_2_2_4 * sqrt(2)
n_tests_2_2_4 <- rowSums(!is.na(tukey_stat_2_2_4))
suppressWarnings(
tukey_pval_2_2_4 <- tukey_stat_2_2_4 %>%
dplyr::mutate(
P_value_A_Infection_high_vs_Infection_low = ptukey(
abs(stat_ih_il),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_high = ptukey(
abs(stat_ih_mh),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_high_vs_Mock_low = ptukey(
abs(stat_ih_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_high = ptukey(
abs(stat_il_mh),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Infection_low_vs_Mock_low = ptukey(
abs(stat_il_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
),
P_value_A_Mock_high_vs_Mock_low = ptukey(
abs(stat_mh_ml),
nranges = 1,
nmeans = n_tests_2_2_4,
df = nona_counts_2_2_4 - n_tests_2_2_4,
lower.tail = FALSE
)
) %>%
dplyr::select(P_value_A_Infection_high_vs_Infection_low,
P_value_A_Infection_high_vs_Mock_high,
P_value_A_Infection_high_vs_Mock_low,
P_value_A_Infection_low_vs_Mock_high,
P_value_A_Infection_low_vs_Mock_low,
P_value_A_Mock_high_vs_Mock_low,) %>%
`rownames<-`(NULL) %>%
data.matrix()
)
# Find rows where just one test occurred.
just_one_2_2_4 <- which(rowSums(!is.na(tukey_stat_2_2_4)) == 1)
# Nab index where each test occurred in the subsetted p-value data frame.
not_na_2_2_4 <- which(!is.na(pval_a_2_2_4[just_one_2_2_4, ]))
# Replace NAs (from only one test being performed) with the original p-value.
tukey_pval_2_2_4[just_one_2_2_4, ][not_na_2_2_4] <- data.matrix(
pval_a_2_2_4[just_one_2_2_4, ]
)[not_na_2_2_4]
# Convert back to a data frame.
tukey_pval_2_2_4 <- data.frame(tukey_pval_2_2_4)
# Set a seed because the mvtnorm::pmvt function has a random process.
set.seed(4)
suppressWarnings(
dunnett_2_2_4 <- pval_a_2_2_4 %>%
dplyr::rowwise() %>%
dplyr::mutate(
no_nas = sum(!is.na(dplyr::c_across(dplyr::everything())))
) %>%
dplyr::ungroup() %>%
dplyr::mutate(dFree = nona_counts_2_2_4 - no_nas) %>%
cbind(test_stat_2_2_4) %>%
dplyr::rowwise() %>%
dplyr::mutate(
pval_ih_il = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_il),
no_nas),
rep(abs(stat_ih_il),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_mh),
no_nas),
rep(abs(stat_ih_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_ih_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_ih_ml),
no_nas),
rep(abs(stat_ih_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_mh = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_mh),
no_nas),
rep(abs(stat_il_mh),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_il_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_il_ml),
no_nas),
rep(abs(stat_il_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
),
pval_mh_ml = tryCatch (
1 - mvtnorm::pmvt(
-rep(abs(stat_mh_ml),
no_nas),
rep(abs(stat_mh_ml),
no_nas),
df = dFree
),
error = function (e) {NA}
)
) %>%
dplyr::select(pval_ih_il, pval_ih_mh, pval_ih_ml,
pval_il_mh, pval_il_ml, pval_mh_ml)
)
# Generate G-Test standards ----------------------------------------------------
# Construct count matrices: Objects that start with obs represent counts of
# non-missing (observed) values and objects that start with abs represent
# counts of missing (absent) values.
obs_inf_1_0_2 <- rowSums(!is.na(gfilta_1_0_2$e_data[, 2:10]))
obs_mock_1_0_2 <- rowSums(!is.na(gfilta_1_0_2$e_data[, 11:13]))
abs_inf_1_0_2 <- rowSums(is.na(gfilta_1_0_2$e_data[, 2:10]))
abs_mock_1_0_2 <- rowSums(is.na(gfilta_1_0_2$e_data[, 11:13]))
obs_mut_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(2, 4, 5, 10)]))
obs_zom_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(3, 6:9)]))
obs_hum_1_1_3 <- rowSums(!is.na(gfilta_1_1_3$e_data[, c(11:13)]))
abs_mut_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(2, 4, 5, 10)]))
abs_zom_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(3, 6:9)]))
abs_hum_1_1_3 <- rowSums(is.na(gfilta_1_1_3$e_data[, c(11:13)]))
obs_mut_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(2, 4, 5, 10)]))
obs_zom_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(3, 6:9)]))
obs_hum_1_2_3 <- rowSums(!is.na(gfilta_1_2_3$e_data[, c(11:13)]))
abs_mut_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(2, 4, 5, 10)]))
abs_zom_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(3, 6:9)]))
abs_hum_1_2_3 <- rowSums(is.na(gfilta_1_2_3$e_data[, c(11:13)]))
obs_inf_high_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, c(2:4, 7:8)]))
obs_inf_low_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, c(5:6, 9:10)]))
obs_mock_2_0_3 <- rowSums(!is.na(gfilta_2_0_3$e_data[, 11:13]))
abs_inf_high_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, c(2:4, 7:8)]))
abs_inf_low_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, c(5:6, 9:10)]))
abs_mock_2_0_3 <- rowSums(is.na(gfilta_2_0_3$e_data[, 11:13]))
obs_ih_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(2, 4, 6)]))
obs_il_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(3, 5, 7)]))
obs_mh_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, c(8, 9, 13)]))
obs_ml_2_1_4 <- rowSums(!is.na(gfilta_2_1_4$e_data[, 10:12]))
abs_ih_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(2, 4, 6)]))
abs_il_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(3, 5, 7)]))
abs_mh_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, c(8, 9, 13)]))
abs_ml_2_1_4 <- rowSums(is.na(gfilta_2_1_4$e_data[, 10:12]))
obs_ih_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(2, 4, 6)]))
obs_il_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(3, 5, 7)]))
obs_mh_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, c(8, 9, 13)]))
obs_ml_2_2_4 <- rowSums(!is.na(gfilta_2_2_4$e_data[, 10:12]))
abs_ih_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(2, 4, 6)]))
abs_il_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(3, 5, 7)]))
abs_mh_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, c(8, 9, 13)]))
abs_ml_2_2_4 <- rowSums(is.na(gfilta_2_2_4$e_data[, 10:12]))
# main effects: 1; covariates: 0; groups: 2 ---------------
pval_g_1_0_2 <- data.frame(
Infection_vs_Mock = pchisq(
q = g(obs1 = obs_inf_1_0_2,
obs2 = obs_mock_1_0_2,
abs1 = abs_inf_1_0_2,
abs2 = abs_mock_1_0_2,
n1 = 9,
n2 = 3,
n_total = 9 + 3),
df = 1,
lower.tail = FALSE
)
)
flag_g_1_0_2 <- data.frame(
Infection_vs_Mock = gflag(
obs1 = obs_inf_1_0_2,
obs2 = obs_mock_1_0_2,
abs1 = abs_inf_1_0_2,
abs2 = abs_mock_1_0_2,
pvals = pval_g_1_0_2[, 1],
cutoff = 0.05
)
)
mean_1_0_2 <- data.frame(
Mean_Infection = rowMeans(gfilta_1_0_2$e_data[, 2:10],
na.rm = TRUE),
Mean_Mock = rowMeans(gfilta_1_0_2$e_data[, 11:13],
na.rm = TRUE)
)
gstan_1_0_2 <- data.frame(
Mass_Tag_ID = gfilta_1_0_2$e_data$Mass_Tag_ID,
Count_Infection = unname(obs_inf_1_0_2),
Count_Mock = unname(obs_mock_1_0_2),
mean_1_0_2,
Fold_change_Infection_vs_Mock = (
mean_1_0_2[, 1] - mean_1_0_2[, 2]
),
P_value_G_Infection_vs_Mock = pval_g_1_0_2[, 1],
Flag_G_Infection_vs_Mock = flag_g_1_0_2[, 1],
row.names = NULL
)
class(gstan_1_0_2) <- c("statRes", "data.frame")
attr(gstan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(gstan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(gstan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_total = c(length(which(flag_g_1_0_2[, 1] == -1))),
Up_anova = c(0),
Down_anova = c(0),
Up_gtest = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_gtest = c(length(which(flag_g_1_0_2[, 1] == -1))),
row.names = NULL
)
attr(gstan_1_0_2, "statistical_test") <- "gtest"
attr(gstan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_0_2, "pval_thresh") <- 0.05
attr(gstan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_0_2$e_data[, -1])) /
prod(dim(gfilta_1_0_2$e_data[, -1]))),
num_samps = dim(gfilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(gstan_1_0_2, "which_X") <- rep(0, nrow(gstan_1_0_2))
attr(gstan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
pval_g_1_1_3 <- data.frame(
P_value_G_mutant_vs_zombie = rep(0, nrow(gfilta_1_1_3$e_data)),
P_value_G_mutant_vs_human = rep(0, nrow(gfilta_1_1_3$e_data)),
P_value_G_zombie_vs_human = rep(0, nrow(gfilta_1_1_3$e_data))
)
for (e in 1:nrow(gfilta_1_1_3$e_data)) {
pval_g_1_1_3[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, 2:10])) ~
attr(groupDF_1_1_3, "covariates")$Gender[1:9] +
groupDF_1_1_3$Group[1:9],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
pval_g_1_1_3[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, c(2, 4, 5, 10:13)])) ~
attr(groupDF_1_1_3, "covariates")$Gender[c(1, 3, 4, 9:12)] +
groupDF_1_1_3$Group[c(1, 3, 4, 9:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
pval_g_1_1_3[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_1_1_3$e_data[e, c(3, 6:9, 11:13)])) ~
attr(groupDF_1_1_3, "covariates")$Gender[c(2, 5:8, 10:12)] +
groupDF_1_1_3$Group[c(2, 5:8, 10:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
}
flag_g_1_1_3 <- data.frame(
Flag_G_mutant_vs_zombie = gflag(
obs1 = obs_mut_1_1_3,
obs2 = obs_zom_1_1_3,
abs1 = abs_mut_1_1_3,
abs2 = abs_zom_1_1_3,
pvals = pval_g_1_1_3[, 1],
cutoff = 0.05
),
Flag_G_mutant_vs_human = gflag(
obs1 = obs_mut_1_1_3,
obs2 = obs_hum_1_1_3,
abs1 = abs_mut_1_1_3,
abs2 = abs_hum_1_1_3,
pvals = pval_g_1_1_3[, 2],
cutoff = 0.05
),
Flag_G_zombie_vs_human = gflag(
obs1 = obs_zom_1_1_3,
obs2 = obs_hum_1_1_3,
abs1 = abs_zom_1_1_3,
abs2 = abs_hum_1_1_3,
pvals = pval_g_1_1_3[, 3],
cutoff = 0.05
)
)
data_1_1_3 <- gfilta_1_1_3$e_data[, -1]
Betas = compute_betas(data_mat = data.matrix(data_1_1_3), Xmatrix = data.matrix(Xmatrix_1_1_3))
group_sampnames <- gfilta_1_1_3$f_data$SampleID
groupData <- groupDF_1_1_3[group_sampnames %in% colnames(gfilta_1_1_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(gfilta_1_1_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", "Gender")
mean_1_1_3 <- get_lsmeans(data = data_1_1_3, xmatrix = Xmatrix_1_1_3, pred_grid = pred_grid, Betas = Betas)
counts_1_1_3 <- data.frame(
"Count_mutant" = unname(obs_mut_1_1_3),
"Count_zombie" = unname(obs_zom_1_1_3),
"Count_human" = unname(obs_hum_1_1_3)
)
mean_1_1_3[counts_1_1_3 == 0] <- NA
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
diffs_1_1_3 <- fold_change_diff(data.matrix(mean_1_1_3), cmat)
gstan_1_1_3 <- data.frame(
Mass_Tag_ID = gfilta_1_1_3$e_data$Mass_Tag_ID,
counts_1_1_3,
mean_1_1_3,
Fold_change_mutant_vs_zombie = diffs_1_1_3[,1],
Fold_change_mutant_vs_human = diffs_1_1_3[,2],
Fold_change_zombie_vs_human = diffs_1_1_3[,3],
pval_g_1_1_3,
flag_g_1_1_3,
row.names = NULL
)
class(gstan_1_1_3) <- c("statRes", "data.frame")
attr(gstan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(gstan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(gstan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_total = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_1_1_3, "statistical_test") <- "gtest"
attr(gstan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_1_3, "pval_thresh") <- 0.05
attr(gstan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_1_3$e_data[, -1])) /
prod(dim(gfilta_1_1_3$e_data[, -1]))),
num_samps = dim(gfilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(gstan_1_1_3, "which_X") <- rep(0, nrow(gstan_1_1_3))
attr(gstan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_1_3, "data_class") <- "pepData"
# main effects: 1; covariates: 2; groups: 3 ---------------
pval_g_1_2_3 <- data.frame(
P_value_G_mutant_vs_zombie = rep(0, nrow(gfilta_1_2_3$e_data)),
P_value_G_mutant_vs_human = rep(0, nrow(gfilta_1_2_3$e_data)),
P_value_G_zombie_vs_human = rep(0, nrow(gfilta_1_2_3$e_data))
)
for (e in 1:nrow(gfilta_1_2_3$e_data)) {
pval_g_1_2_3[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, 2:10])) ~
attr(groupDF_1_2_3, "covariates")$Gender[1:9] +
attr(groupDF_1_2_3, "covariates")$Age[1:9] +
groupDF_1_2_3$Group[1:9],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
pval_g_1_2_3[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, c(2, 4, 5, 10:13)])) ~
attr(groupDF_1_2_3, "covariates")$Gender[c(1, 3, 4, 9:12)] +
attr(groupDF_1_2_3, "covariates")$Age[c(1, 3, 4, 9:12)] +
groupDF_1_2_3$Group[c(1, 3, 4, 9:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
pval_g_1_2_3[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_1_2_3$e_data[e, c(3, 6:9, 11:13)])) ~
attr(groupDF_1_2_3, "covariates")$Gender[c(2, 5:8, 10:12)] +
attr(groupDF_1_2_3, "covariates")$Age[c(2, 5:8, 10:12)] +
groupDF_1_2_3$Group[c(2, 5:8, 10:12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
}
flag_g_1_2_3 <- data.frame(
Flag_G_mutant_vs_zombie = gflag(
obs1 = obs_mut_1_2_3,
obs2 = obs_zom_1_2_3,
abs1 = abs_mut_1_2_3,
abs2 = abs_zom_1_2_3,
pvals = pval_g_1_2_3[, 1],
cutoff = 0.05
),
Flag_G_mutant_vs_human = gflag(
obs1 = obs_mut_1_2_3,
obs2 = obs_hum_1_2_3,
abs1 = abs_mut_1_2_3,
abs2 = abs_hum_1_2_3,
pvals = pval_g_1_2_3[, 2],
cutoff = 0.05
),
Flag_G_zombie_vs_human = gflag(
obs1 = obs_zom_1_2_3,
obs2 = obs_hum_1_2_3,
abs1 = abs_zom_1_2_3,
abs2 = abs_hum_1_2_3,
pvals = pval_g_1_2_3[, 3],
cutoff = 0.05
)
)
data_1_2_3 = data.matrix(gfilta_1_2_3$e_data[, -1])
Betas = compute_betas(data_mat = data_1_2_3, Xmatrix = data.matrix(Xmatrix_1_2_3))
group_sampnames <- gfilta_1_2_3$f_data$SampleID
groupData <- groupDF_1_2_3[group_sampnames %in% colnames(gfilta_1_2_3$e_data),]
groupData <- groupData %>%
dplyr::left_join(gfilta_1_2_3$f_data)
groupData[,"Condition"] <- lapply(groupData["Condition"], function(x) factor(x, levels=unique(x)))
pred_grid <- get_pred_grid(groupData, "Condition", c("Gender", "Age"))
mean_1_2_3 <- get_lsmeans(data = data_1_2_3, xmatrix = Xmatrix_1_2_3, pred_grid = pred_grid, Betas = Betas, continuous_covar_inds = 5)
counts_1_2_3 <- data.frame(
"Count_mutant" = unname(obs_mut_1_2_3),
"Count_zombie" = unname(obs_zom_1_2_3),
"Count_human" = unname(obs_hum_1_2_3)
)
mean_1_2_3[counts_1_2_3 == 0] <- NA
cmat = rbind(c(1, -1, 0), c(1, 0, -1), c(0, 1, -1))
diffs_1_2_3 <- fold_change_diff(data.matrix(mean_1_2_3), cmat)
gstan_1_2_3 <- data.frame(
Mass_Tag_ID = gfilta_1_2_3$e_data$Mass_Tag_ID,
counts_1_2_3,
mean_1_2_3,
Fold_change_mutant_vs_zombie = diffs_1_2_3[,1],
Fold_change_mutant_vs_human = diffs_1_2_3[,2],
Fold_change_zombie_vs_human = diffs_1_2_3[,3],
pval_g_1_2_3,
flag_g_1_2_3,
row.names = NULL
)
class(gstan_1_2_3) <- c("statRes", "data.frame")
attr(gstan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(gstan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(gstan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_total = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_1_2_3, "statistical_test") <- "gtest"
attr(gstan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_1_2_3, "pval_thresh") <- 0.05
attr(gstan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_2_3$e_data[, -1])) /
prod(dim(gfilta_1_2_3$e_data[, -1]))),
num_samps = dim(gfilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(gstan_1_2_3, "which_X") <- rep(0, nrow(gstan_1_2_3))
attr(gstan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_1_2_3, "data_class") <- "pepData"
# main effects: 2; covariates: 0; groups: 3 ---------------
pval_g_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = pchisq(
q = g(obs1 = obs_inf_high_2_0_3,
obs2 = obs_inf_low_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_inf_low_2_0_3,
n1 = 5,
n2 = 4,
n_total = 5 + 4),
df = 1,
lower.tail = FALSE
),
Infection_high_vs_Mock_none = pchisq(
q = g(obs1 = obs_inf_high_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_mock_2_0_3,
n1 = 5,
n2 = 3,
n_total = 5 + 3),
df = 1,
lower.tail = FALSE
),
Infection_low_vs_Mock_none = pchisq(
q = g(obs1 = obs_inf_low_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_low_2_0_3,
abs2 = abs_mock_2_0_3,
n1 = 4,
n2 = 3,
n_total = 4 + 3),
df = 1,
lower.tail = FALSE
)
)
flag_g_2_0_3 <- data.frame(
Infection_high_vs_Infection_low = gflag(
obs1 = obs_inf_high_2_0_3,
obs2 = obs_inf_low_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_inf_low_2_0_3,
pvals = pval_g_2_0_3[, 1],
cutoff = 0.05
),
Infection_high_vs_Mock_none = gflag(
obs1 = obs_inf_high_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_high_2_0_3,
abs2 = abs_mock_2_0_3,
pvals = pval_g_2_0_3[, 2],
cutoff = 0.05
),
Infection_low_vs_Mock_none = gflag(
obs1 = obs_inf_low_2_0_3,
obs2 = obs_mock_2_0_3,
abs1 = abs_inf_low_2_0_3,
abs2 = abs_mock_2_0_3,
pvals = pval_g_2_0_3[, 3],
cutoff = 0.05
)
)
mean_2_0_3 <- data.frame(
Mean_Infection_high = rowMeans(gfilta_2_0_3$e_data[, c(2:4, 7:8)],
na.rm = TRUE),
Mean_Infection_low = rowMeans(gfilta_2_0_3$e_data[, c(5:6, 9:10)],
na.rm = TRUE),
Mean_Mock_none = rowMeans(gfilta_2_0_3$e_data[, 11:13],
na.rm = TRUE)
)
gstan_2_0_3 <- data.frame(
Mass_Tag_ID = gfilta_2_0_3$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_inf_high_2_0_3),
Count_Infection_low = unname(obs_inf_low_2_0_3),
Count_Mock_none = unname(obs_mock_2_0_3),
mean_2_0_3,
Fold_change_Infection_high_vs_Infection_low = (
mean_2_0_3[, 1] - mean_2_0_3[, 2]
),
Fold_change_Infection_high_vs_Mock_none = (
mean_2_0_3[, 1] - mean_2_0_3[, 3]
),
Fold_change_Infection_low_vs_Mock_none = (
mean_2_0_3[, 2] - mean_2_0_3[, 3]
),
P_value_G_Infection_high_vs_Infection_low = pval_g_2_0_3[, 1],
P_value_G_Infection_high_vs_Mock_none = pval_g_2_0_3[, 2],
P_value_G_Infection_low_vs_Mock_none = pval_g_2_0_3[, 3],
Flag_G_Infection_high_vs_Infection_low = flag_g_2_0_3[, 1],
Flag_G_Infection_high_vs_Mock_none = flag_g_2_0_3[, 2],
Flag_G_Infection_low_vs_Mock_none = flag_g_2_0_3[, 3],
row.names = NULL
)
class(gstan_2_0_3) <- c("statRes", "data.frame")
attr(gstan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(gstan_2_0_3, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none")
attr(gstan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_total = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
Up_anova = c(0, 0, 0),
Down_anova = c(0, 0, 0),
Up_gtest = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
row.names = NULL
)
attr(gstan_2_0_3, "statistical_test") <- "gtest"
attr(gstan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_0_3, "pval_thresh") <- 0.05
attr(gstan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_0_3$e_data[, -1])) /
prod(dim(gfilta_2_0_3$e_data[, -1]))),
num_samps = dim(gfilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(gstan_2_0_3, "which_X") <- rep(0, nrow(gstan_2_0_3))
attr(gstan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_0_3, "data_class") <- "pepData"
# main effects: 2; covariates: 1; groups: 4 ---------------
pval_g_2_1_4 <- data.frame(
P_value_G_Infection_high_vs_Infection_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_high_vs_Mock_high = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_high_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_low_vs_Mock_high = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Infection_low_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data)),
P_value_G_Mock_high_vs_Mock_low = rep(0, nrow(gfilta_2_1_4$e_data))
)
for (e in 1:nrow(gfilta_2_1_4$e_data)) {
# Compute the p-value for the Infection_high_vs_Infection_low test.
pval_g_2_1_4[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, 2:7])) ~
attr(groupDF_2_1_4, "covariates")$Gender[1:6] +
groupDF_2_1_4$Group[1:6],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_high_vs_Mock_high test.
pval_g_2_1_4[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(2, 4, 6, 8, 9, 13)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(1, 3, 5, 7, 8, 12)] +
groupDF_2_1_4$Group[c(1, 3, 5, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_high_vs_Mock_low test.
pval_g_2_1_4[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(2, 4, 6, 10:12)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(1, 3, 5, 9:11)] +
groupDF_2_1_4$Group[c(1, 3, 5, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_low_vs_Mock_high test.
pval_g_2_1_4[e, 4] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(3, 5, 7, 8, 9, 13)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(2, 4, 6, 7, 8, 12)] +
groupDF_2_1_4$Group[c(2, 4, 6, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Infection_low_vs_Mock_low test.
pval_g_2_1_4[e, 5] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, c(3, 5, 7, 10:12)])) ~
attr(groupDF_2_1_4, "covariates")$Gender[c(2, 4, 6, 9:11)] +
groupDF_2_1_4$Group[c(2, 4, 6, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
# Compute the p-value for the Mock_high_vs_Mock_low test.
pval_g_2_1_4[e, 6] <- anova(
glm(
as.numeric(!is.na(gfilta_2_1_4$e_data[e, 8:13])) ~
attr(groupDF_2_1_4, "covariates")$Gender[7:12] +
groupDF_2_1_4$Group[7:12],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[3]]
}
flag_g_2_1_4 <- data.frame(
Flag_G_Infection_high_vs_Infection_low = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_il_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_il_2_1_4,
pvals = pval_g_2_1_4[, 1],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_high = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_mh_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_mh_2_1_4,
pvals = pval_g_2_1_4[, 2],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_low = gflag(
obs1 = obs_ih_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_ih_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 3],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_high = gflag(
obs1 = obs_il_2_1_4,
obs2 = obs_mh_2_1_4,
abs1 = abs_il_2_1_4,
abs2 = abs_mh_2_1_4,
pvals = pval_g_2_1_4[, 4],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_low = gflag(
obs1 = obs_il_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_il_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 5],
cutoff = 0.05
),
Flag_G_Mock_high_vs_Mock_low = gflag(
obs1 = obs_mh_2_1_4,
obs2 = obs_ml_2_1_4,
abs1 = abs_mh_2_1_4,
abs2 = abs_ml_2_1_4,
pvals = pval_g_2_1_4[, 6],
cutoff = 0.05
)
)
gdf = dplyr::left_join(groupDF_2_1_4, attr(groupDF_2_1_4, "covariates")) %>%
dplyr::select(-SampleID)
dragon <- run_twofactor_cpp(
data = data.matrix(gfilta_2_1_4$e_data[, -1]),
gpData = gdf,
Xfull = Xmatrix_2_1_4_full,
Xred = Xmatrix_2_1_4,
pred_grid_full = pred_grid_full_2_1_4,
pred_grid_red = pred_grid_red_2_1_4,
continuous_covar_inds = numeric(0)
)
mean_2_1_4 <- data.frame(
Mean_Infection_high = dragon$adj_group_means[, 1],
Mean_Infection_low = dragon$adj_group_means[, 2],
Mean_Mock_high = dragon$adj_group_means[, 3],
Mean_Mock_low = dragon$adj_group_means[, 4]
)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
diffs_2_1_4 <- fold_change_diff(data.matrix(mean_2_1_4), cmat)
diffs_2_1_4 <- as.data.frame(diffs_2_1_4) %>%
`colnames<-`(c("Fold_change_Infection_high_vs_Infection_low", "Fold_change_Infection_high_vs_Mock_high", "Fold_change_Infection_high_vs_Mock_low", "Fold_change_Infection_low_vs_Mock_high", "Fold_change_Infection_low_vs_Mock_low", "Fold_change_Mock_high_vs_Mock_low"))
gstan_2_1_4 <- data.frame(
Mass_Tag_ID = gfilta_2_1_4$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_ih_2_1_4),
Count_Infection_low = unname(obs_il_2_1_4),
Count_Mock_high = unname(obs_mh_2_1_4),
Count_Mock_low = unname(obs_ml_2_1_4),
mean_2_1_4,
diffs_2_1_4,
pval_g_2_1_4,
flag_g_2_1_4,
row.names = NULL
)
class(gstan_2_1_4) <- c("statRes", "data.frame")
attr(gstan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(gstan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(gstan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_total = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
Up_anova = c(0, 0, 0, 0, 0, 0),
Down_anova = c(0, 0, 0, 0, 0, 0),
Up_gtest = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
row.names = NULL
)
attr(gstan_2_1_4, "statistical_test") <- "gtest"
attr(gstan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_1_4, "pval_thresh") <- 0.05
attr(gstan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_1_4$e_data[, -1])) /
prod(dim(gfilta_2_1_4$e_data[, -1]))),
num_samps = dim(gfilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(gstan_2_1_4, "which_X") <- which_X_2_1_4_g
attr(gstan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_1_4, "data_class") <- "pepData"
# main effects: 2; covariates: 2; groups: 4 ---------------
pval_g_2_2_4 <- data.frame(
P_value_G_Infection_high_vs_Infection_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_high_vs_Mock_high = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_high_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_low_vs_Mock_high = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Infection_low_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data)),
P_value_G_Mock_high_vs_Mock_low = rep(0, nrow(gfilta_2_2_4$e_data))
)
for (e in 1:nrow(gfilta_2_2_4$e_data)) {
# Compute the p-value for the Infection_high_vs_Infection_low test.
pval_g_2_2_4[e, 1] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, 2:7])) ~
attr(groupDF_2_2_4, "covariates")$Gender[1:6] +
attr(groupDF_2_2_4, "covariates")$Age[1:6] +
groupDF_2_2_4$Group[1:6],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_high_vs_Mock_high test.
pval_g_2_2_4[e, 2] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(2, 4, 6, 8, 9, 13)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(1, 3, 5, 7, 8, 12)] +
attr(groupDF_2_2_4, "covariates")$Age[c(1, 3, 5, 7, 8, 12)] +
groupDF_2_2_4$Group[c(1, 3, 5, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_high_vs_Mock_low test.
pval_g_2_2_4[e, 3] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(2, 4, 6, 10:12)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(1, 3, 5, 9:11)] +
attr(groupDF_2_2_4, "covariates")$Age[c(1, 3, 5, 9:11)] +
groupDF_2_2_4$Group[c(1, 3, 5, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_low_vs_Mock_high test.
pval_g_2_2_4[e, 4] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(3, 5, 7, 8, 9, 13)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(2, 4, 6, 7, 8, 12)] +
attr(groupDF_2_2_4, "covariates")$Age[c(2, 4, 6, 7, 8, 12)] +
groupDF_2_2_4$Group[c(2, 4, 6, 7, 8, 12)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Infection_low_vs_Mock_low test.
pval_g_2_2_4[e, 5] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, c(3, 5, 7, 10:12)])) ~
attr(groupDF_2_2_4, "covariates")$Gender[c(2, 4, 6, 9:11)] +
attr(groupDF_2_2_4, "covariates")$Age[c(2, 4, 6, 9:11)] +
groupDF_2_2_4$Group[c(2, 4, 6, 9:11)],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
# Compute the p-value for the Mock_high_vs_Mock_low test.
pval_g_2_2_4[e, 6] <- anova(
glm(
as.numeric(!is.na(gfilta_2_2_4$e_data[e, 8:13])) ~
attr(groupDF_2_2_4, "covariates")$Gender[7:12] +
attr(groupDF_2_2_4, "covariates")$Age[7:12] +
groupDF_2_2_4$Group[7:12],
family = binomial
),
test = "Chisq"
)$`Pr(>Chi)`[[4]]
}
flag_g_2_2_4 <- data.frame(
Flag_G_Infection_high_vs_Infection_low = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_il_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_il_2_2_4,
pvals = pval_g_2_2_4[, 1],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_high = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_mh_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_mh_2_2_4,
pvals = pval_g_2_2_4[, 2],
cutoff = 0.05
),
Flag_G_Infection_high_vs_Mock_low = gflag(
obs1 = obs_ih_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_ih_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 3],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_high = gflag(
obs1 = obs_il_2_2_4,
obs2 = obs_mh_2_2_4,
abs1 = abs_il_2_2_4,
abs2 = abs_mh_2_2_4,
pvals = pval_g_2_2_4[, 4],
cutoff = 0.05
),
Flag_G_Infection_low_vs_Mock_low = gflag(
obs1 = obs_il_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_il_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 5],
cutoff = 0.05
),
Flag_G_Mock_high_vs_Mock_low = gflag(
obs1 = obs_mh_2_2_4,
obs2 = obs_ml_2_2_4,
abs1 = abs_mh_2_2_4,
abs2 = abs_ml_2_2_4,
pvals = pval_g_2_2_4[, 6],
cutoff = 0.05
)
)
gdf = dplyr::left_join(groupDF_2_2_4, attr(groupDF_2_2_4, "covariates")) %>%
dplyr::select(-SampleID)
mustang <- run_twofactor_cpp(
data = data.matrix(gfilta_2_2_4$e_data[, -1]),
gpData = gdf,
Xfull=Xmatrix_2_2_4_full, Xred = Xmatrix_2_2_4,
pred_grid_full = pred_grid_full_2_2_4, pred_grid_red = pred_grid_red_2_2_4,
continuous_covar_inds = 5
)
mean_2_2_4 <- data.frame(
Mean_Infection_high = mustang$adj_group_means[, 1],
Mean_Infection_low = mustang$adj_group_means[, 2],
Mean_Mock_high = mustang$adj_group_means[, 3],
Mean_Mock_low = mustang$adj_group_means[, 4]
)
cmat = rbind(c(1, -1, 0, 0), c(1, 0, -1, 0), c(1, 0, 0, -1),
c(0, 1, -1, 0), c(0, 1, 0, -1), c(0, 0, 1, -1))
diffs_2_2_4 <- fold_change_diff(data.matrix(mean_2_2_4), cmat)
diffs_2_2_4 <- as.data.frame(diffs_2_2_4) %>%
`colnames<-`(c("Fold_change_Infection_high_vs_Infection_low", "Fold_change_Infection_high_vs_Mock_high", "Fold_change_Infection_high_vs_Mock_low", "Fold_change_Infection_low_vs_Mock_high", "Fold_change_Infection_low_vs_Mock_low", "Fold_change_Mock_high_vs_Mock_low"))
gstan_2_2_4 <- data.frame(
Mass_Tag_ID = gfilta_2_2_4$e_data$Mass_Tag_ID,
Count_Infection_high = unname(obs_ih_2_2_4),
Count_Infection_low = unname(obs_il_2_2_4),
Count_Mock_high = unname(obs_mh_2_2_4),
Count_Mock_low = unname(obs_ml_2_2_4),
mean_2_2_4,
diffs_2_2_4,
pval_g_2_2_4,
flag_g_2_2_4,
row.names = NULL
)
class(gstan_2_2_4) <- c("statRes", "data.frame")
attr(gstan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(gstan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(gstan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_total = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
Up_anova = c(0, 0, 0, 0, 0, 0),
Down_anova = c(0, 0, 0, 0, 0, 0),
Up_gtest = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
row.names = NULL
)
attr(gstan_2_2_4, "statistical_test") <- "gtest"
attr(gstan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(gstan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(gstan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(gstan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(gstan_2_2_4, "pval_thresh") <- 0.05
attr(gstan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_2_4$e_data[, -1])) /
prod(dim(gfilta_2_2_4$e_data[, -1]))),
num_samps = dim(gfilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(gstan_2_2_4, "which_X") <- which_X_2_2_4_g
attr(gstan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(gstan_2_2_4, "data_class") <- "pepData"
# Create combined standards ----------------------------------------------------
# main effects: 1; covariates: 0; groups: 2 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_0_2 <- dplyr::full_join(gstan_1_0_2[, c(1:3, 7:8)],
astan_1_0_2) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_0_2[is.nan(data.matrix(cstan_1_0_2))] <- NA
class(cstan_1_0_2) <- c("statRes", "data.frame")
attr(cstan_1_0_2, "group_DF") <- groupDF_1_0_2
attr(cstan_1_0_2, "comparisons") <- c("Infection_vs_Mock")
attr(cstan_1_0_2, "number_significant") <- data.frame(
Comparison = c("Infection_vs_Mock"),
Up_total = sum(length(which(flag_g_1_0_2[, 1] == 1)),
length(which(flag_a_1_0_2[, 1] == 1))),
Down_total = sum(length(which(flag_g_1_0_2[, 1] == -1)),
length(which(flag_a_1_0_2[, 1] == -1))),
Up_anova = c(length(which(flag_a_1_0_2[, 1] == 1))),
Down_anova = c(length(which(flag_a_1_0_2[, 1] == -1))),
Up_gtest = c(length(which(flag_g_1_0_2[, 1] == 1))),
Down_gtest = c(length(which(flag_g_1_0_2[, 1] == -1))),
row.names = NULL
)
attr(cstan_1_0_2, "statistical_test") <- "combined"
attr(cstan_1_0_2, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_0_2, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_0_2, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_0_2, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_0_2, "pval_thresh") <- 0.05
attr(cstan_1_0_2, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_0_2$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_0_2$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_0_2$e_data[, -1])) /
prod(dim(gfilta_1_0_2$e_data[, -1]))),
num_samps = dim(gfilta_1_0_2$f_data)[1],
data_types = NULL
)
attr(cstan_1_0_2, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, NA,
0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, NA,
0, 0, NA, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0,
0, 0, NA, NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0,
0, 0, NA, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA,
0, 0)
attr(cstan_1_0_2, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_0_2$e_data$Mass_Tag_ID,
Infection_vs_Mock = dplyr::case_when(
is.na(cstan_1_0_2$Flag_A_Infection_vs_Mock) ~
cstan_1_0_2$Flag_G_Infection_vs_Mock,
(cstan_1_0_2$P_value_A_Infection_vs_Mock > 0.05 &
cstan_1_0_2$P_value_G_Infection_vs_Mock < 0.05) ~
cstan_1_0_2$Flag_G_Infection_vs_Mock,
!is.na(cstan_1_0_2$Flag_A_Infection_vs_Mock) ~
cstan_1_0_2$Flag_A_Infection_vs_Mock
)
)
attr(cstan_1_0_2, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_0_2, "data_class") <- "pepData"
# main effects: 1; covariates: 1; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_1_3 <- dplyr::full_join(gstan_1_1_3[, c(1:4, 11:16)],
astan_1_1_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_1_3[is.nan(data.matrix(cstan_1_1_3))] <- NA
class(cstan_1_1_3) <- c("statRes", "data.frame")
attr(cstan_1_1_3, "group_DF") <- groupDF_1_1_3
attr(cstan_1_1_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(cstan_1_1_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(sum(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 1] == 1))),
sum(length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 2] == 1))),
sum(length(which(flag_g_1_1_3[, 3] == 1)),
length(which(flag_a_1_1_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 1] == -1))),
sum(length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 2] == -1))),
sum(length(which(flag_g_1_1_3[, 3] == -1)),
length(which(flag_a_1_1_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_1_1_3[, 1] == 1)),
length(which(flag_a_1_1_3[, 2] == 1)),
length(which(flag_a_1_1_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_1_3[, 1] == -1)),
length(which(flag_a_1_1_3[, 2] == -1)),
length(which(flag_a_1_1_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_1_1_3[, 1] == 1)),
length(which(flag_g_1_1_3[, 2] == 1)),
length(which(flag_g_1_1_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_1_3[, 1] == -1)),
length(which(flag_g_1_1_3[, 2] == -1)),
length(which(flag_g_1_1_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_1_1_3, "statistical_test") <- "combined"
attr(cstan_1_1_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_1_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_1_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_1_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_1_3, "pval_thresh") <- 0.05
attr(cstan_1_1_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_1_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_1_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_1_3$e_data[, -1])) /
prod(dim(gfilta_1_1_3$e_data[, -1]))),
num_samps = dim(gfilta_1_1_3$f_data)[1],
data_types = NULL
)
attr(cstan_1_1_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_1_1_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_1_3$e_data$Mass_Tag_ID,
mutant_vs_zombie = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
is.na(cstan_1_1_3$P_value_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
(cstan_1_1_3$P_value_A_mutant_vs_zombie > 0.05 &
cstan_1_1_3$P_value_G_mutant_vs_zombie < 0.05) ~
cstan_1_1_3$Flag_G_mutant_vs_zombie,
!is.na(cstan_1_1_3$Flag_A_mutant_vs_zombie) ~
cstan_1_1_3$Flag_A_mutant_vs_zombie
),
mutant_vs_human = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_mutant_vs_human) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
is.na(cstan_1_1_3$P_value_A_mutant_vs_human) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
(cstan_1_1_3$P_value_A_mutant_vs_human > 0.05 &
cstan_1_1_3$P_value_G_mutant_vs_human < 0.05) ~
cstan_1_1_3$Flag_G_mutant_vs_human,
!is.na(cstan_1_1_3$Flag_A_mutant_vs_human) ~
cstan_1_1_3$Flag_A_mutant_vs_human
),
zombie_vs_human = dplyr::case_when(
is.na(cstan_1_1_3$Flag_A_zombie_vs_human) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
is.na(cstan_1_1_3$P_value_A_zombie_vs_human) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
(cstan_1_1_3$P_value_A_zombie_vs_human > 0.05 &
cstan_1_1_3$P_value_G_zombie_vs_human < 0.05) ~
cstan_1_1_3$Flag_G_zombie_vs_human,
!is.na(cstan_1_1_3$Flag_A_zombie_vs_human) ~
cstan_1_1_3$Flag_A_zombie_vs_human
)
)
attr(cstan_1_1_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_1_3, "data_class") <- "pepData"
# main effects: 1; covariates: 2; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_1_2_3 <- dplyr::full_join(gstan_1_2_3[, c(1:4, 11:16)],
astan_1_2_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_1_2_3[is.nan(data.matrix(cstan_1_2_3))] <- NA
class(cstan_1_2_3) <- c("statRes", "data.frame")
attr(cstan_1_2_3, "group_DF") <- groupDF_1_2_3
attr(cstan_1_2_3, "comparisons") <- c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human")
attr(cstan_1_2_3, "number_significant") <- data.frame(
Comparison = c("mutant_vs_zombie",
"mutant_vs_human",
"zombie_vs_human"),
Up_total = c(sum(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 1] == 1))),
sum(length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 2] == 1))),
sum(length(which(flag_g_1_2_3[, 3] == 1)),
length(which(flag_a_1_2_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 1] == -1))),
sum(length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 2] == -1))),
sum(length(which(flag_g_1_2_3[, 3] == -1)),
length(which(flag_a_1_2_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_1_2_3[, 1] == 1)),
length(which(flag_a_1_2_3[, 2] == 1)),
length(which(flag_a_1_2_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_1_2_3[, 1] == -1)),
length(which(flag_a_1_2_3[, 2] == -1)),
length(which(flag_a_1_2_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_1_2_3[, 1] == 1)),
length(which(flag_g_1_2_3[, 2] == 1)),
length(which(flag_g_1_2_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_1_2_3[, 1] == -1)),
length(which(flag_g_1_2_3[, 2] == -1)),
length(which(flag_g_1_2_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_1_2_3, "statistical_test") <- "combined"
attr(cstan_1_2_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_1_2_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_1_2_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_1_2_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_1_2_3, "pval_thresh") <- 0.05
attr(cstan_1_2_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_1_2_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_1_2_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_1_2_3$e_data[, -1])) /
prod(dim(gfilta_1_2_3$e_data[, -1]))),
num_samps = dim(gfilta_1_2_3$f_data)[1],
data_types = NULL
)
attr(cstan_1_2_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_1_2_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_1_2_3$e_data$Mass_Tag_ID,
mutant_vs_zombie = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
is.na(cstan_1_2_3$P_value_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
(cstan_1_2_3$P_value_A_mutant_vs_zombie > 0.05 &
cstan_1_2_3$P_value_G_mutant_vs_zombie < 0.05) ~
cstan_1_2_3$Flag_G_mutant_vs_zombie,
!is.na(cstan_1_2_3$Flag_A_mutant_vs_zombie) ~
cstan_1_2_3$Flag_A_mutant_vs_zombie
),
mutant_vs_human = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_mutant_vs_human) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
is.na(cstan_1_2_3$P_value_A_mutant_vs_human) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
(cstan_1_2_3$P_value_A_mutant_vs_human > 0.05 &
cstan_1_2_3$P_value_G_mutant_vs_human < 0.05) ~
cstan_1_2_3$Flag_G_mutant_vs_human,
!is.na(cstan_1_2_3$Flag_A_mutant_vs_human) ~
cstan_1_2_3$Flag_A_mutant_vs_human
),
zombie_vs_human = dplyr::case_when(
is.na(cstan_1_2_3$Flag_A_zombie_vs_human) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
is.na(cstan_1_2_3$P_value_A_zombie_vs_human) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
(cstan_1_2_3$P_value_A_zombie_vs_human > 0.05 &
cstan_1_2_3$P_value_G_zombie_vs_human < 0.05) ~
cstan_1_2_3$Flag_G_zombie_vs_human,
!is.na(cstan_1_2_3$Flag_A_zombie_vs_human) ~
cstan_1_2_3$Flag_A_zombie_vs_human
)
)
attr(cstan_1_2_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = "Protein",
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_1_2_3, "data_class") <- "pepData"
# main effects: 2; covariates: 0; groups: 3 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_0_3 <- dplyr::full_join(gstan_2_0_3[, c(1:4, 11:16)],
astan_2_0_3) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_0_3[is.nan(data.matrix(cstan_2_0_3))] <- NA
class(cstan_2_0_3) <- c("statRes", "data.frame")
attr(cstan_2_0_3, "group_DF") <- groupDF_2_0_3
attr(cstan_2_0_3, "comparisons") <- c(
"Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"
)
attr(cstan_2_0_3, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_none",
"Infection_low_vs_Mock_none"),
Up_total = c(sum(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 1] == 1))),
sum(length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 2] == 1))),
sum(length(which(flag_g_2_0_3[, 3] == 1)),
length(which(flag_a_2_0_3[, 3] == 1)))),
Down_total = c(sum(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 1] == -1))),
sum(length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 2] == -1))),
sum(length(which(flag_g_2_0_3[, 3] == -1)),
length(which(flag_a_2_0_3[, 3] == -1)))),
Up_anova = c(length(which(flag_a_2_0_3[, 1] == 1)),
length(which(flag_a_2_0_3[, 2] == 1)),
length(which(flag_a_2_0_3[, 3] == 1))),
Down_anova = c(length(which(flag_a_2_0_3[, 1] == -1)),
length(which(flag_a_2_0_3[, 2] == -1)),
length(which(flag_a_2_0_3[, 3] == -1))),
Up_gtest = c(length(which(flag_g_2_0_3[, 1] == 1)),
length(which(flag_g_2_0_3[, 2] == 1)),
length(which(flag_g_2_0_3[, 3] == 1))),
Down_gtest = c(length(which(flag_g_2_0_3[, 1] == -1)),
length(which(flag_g_2_0_3[, 2] == -1)),
length(which(flag_g_2_0_3[, 3] == -1))),
row.names = NULL
)
attr(cstan_2_0_3, "statistical_test") <- "combined"
attr(cstan_2_0_3, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_0_3, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_0_3, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_0_3, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_0_3, "pval_thresh") <- 0.05
attr(cstan_2_0_3, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_0_3$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_0_3$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_0_3$e_data[, -1])) /
prod(dim(gfilta_2_0_3$e_data[, -1]))),
num_samps = dim(gfilta_2_0_3$f_data)[1],
data_types = NULL
)
attr(cstan_2_0_3, "which_X") <-
c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, NA, NA,
0, 0, 0, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0,
0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, NA, NA, 0, 0, 0, 0,
0, 0, NA, NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_0_3, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_0_3$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_0_3$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_0_3$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_0_3$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_0_3$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_none = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
is.na(cstan_2_0_3$P_value_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
(cstan_2_0_3$P_value_A_Infection_high_vs_Mock_none > 0.05 &
cstan_2_0_3$P_value_G_Infection_high_vs_Mock_none < 0.05) ~
cstan_2_0_3$Flag_G_Infection_high_vs_Mock_none,
!is.na(cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none) ~
cstan_2_0_3$Flag_A_Infection_high_vs_Mock_none
),
Infection_low_vs_Mock_none = dplyr::case_when(
is.na(cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
is.na(cstan_2_0_3$P_value_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
(cstan_2_0_3$P_value_A_Infection_low_vs_Mock_none > 0.05 &
cstan_2_0_3$P_value_G_Infection_low_vs_Mock_none < 0.05) ~
cstan_2_0_3$Flag_G_Infection_low_vs_Mock_none,
!is.na(cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none) ~
cstan_2_0_3$Flag_A_Infection_low_vs_Mock_none
)
)
attr(cstan_2_0_3, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_0_3, "data_class") <- "pepData"
# main effects: 2; covariates: 1; groups: 4 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_1_4 <- dplyr::full_join(gstan_2_1_4[, c(1:5, 16:27)],
astan_2_1_4) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_1_4[is.nan(data.matrix(cstan_2_1_4))] <- NA
class(cstan_2_1_4) <- c("statRes", "data.frame")
attr(cstan_2_1_4, "group_DF") <- groupDF_2_1_4
attr(cstan_2_1_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(cstan_2_1_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(sum(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 1] == 1))),
sum(length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 2] == 1))),
sum(length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 3] == 1))),
sum(length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 4] == 1))),
sum(length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 5] == 1))),
sum(length(which(flag_g_2_1_4[, 6] == 1)),
length(which(flag_a_2_1_4[, 6] == 1)))),
Down_total = c(sum(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 1] == -1))),
sum(length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 2] == -1))),
sum(length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 3] == -1))),
sum(length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 4] == -1))),
sum(length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 5] == -1))),
sum(length(which(flag_g_2_1_4[, 6] == -1)),
length(which(flag_a_2_1_4[, 6] == -1)))),
Up_anova = c(length(which(flag_a_2_1_4[, 1] == 1)),
length(which(flag_a_2_1_4[, 2] == 1)),
length(which(flag_a_2_1_4[, 3] == 1)),
length(which(flag_a_2_1_4[, 4] == 1)),
length(which(flag_a_2_1_4[, 5] == 1)),
length(which(flag_a_2_1_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_1_4[, 1] == -1)),
length(which(flag_a_2_1_4[, 2] == -1)),
length(which(flag_a_2_1_4[, 3] == -1)),
length(which(flag_a_2_1_4[, 4] == -1)),
length(which(flag_a_2_1_4[, 5] == -1)),
length(which(flag_a_2_1_4[, 6] == -1))),
Up_gtest = c(length(which(flag_g_2_1_4[, 1] == 1)),
length(which(flag_g_2_1_4[, 2] == 1)),
length(which(flag_g_2_1_4[, 3] == 1)),
length(which(flag_g_2_1_4[, 4] == 1)),
length(which(flag_g_2_1_4[, 5] == 1)),
length(which(flag_g_2_1_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_1_4[, 1] == -1)),
length(which(flag_g_2_1_4[, 2] == -1)),
length(which(flag_g_2_1_4[, 3] == -1)),
length(which(flag_g_2_1_4[, 4] == -1)),
length(which(flag_g_2_1_4[, 5] == -1)),
length(which(flag_g_2_1_4[, 6] == -1))),
row.names = NULL
)
attr(cstan_2_1_4, "statistical_test") <- "combined"
attr(cstan_2_1_4, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_1_4, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_1_4, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_1_4, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_1_4, "pval_thresh") <- 0.05
attr(cstan_2_1_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_1_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_1_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_1_4$e_data[, -1])) /
prod(dim(gfilta_2_1_4$e_data[, -1]))),
num_samps = dim(gfilta_2_1_4$f_data)[1],
data_types = NULL
)
attr(cstan_2_1_4, "which_X") <-
c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
1, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0, 1,
NA, 0, 0, 0, 1, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_1_4, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_1_4$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_1_4$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_high > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Mock_high < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_high,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Mock_high
),
Infection_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
(cstan_2_1_4$P_value_A_Infection_high_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_high_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_high_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Infection_high_vs_Mock_low
),
Infection_low_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
is.na(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_high > 0.05 &
cstan_2_1_4$P_value_G_Infection_low_vs_Mock_high < 0.05) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_high,
!is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_1_4$Flag_A_Infection_low_vs_Mock_high
),
Infection_low_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
(cstan_2_1_4$P_value_A_Infection_low_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Infection_low_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Infection_low_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Infection_low_vs_Mock_low
),
Mock_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
is.na(cstan_2_1_4$P_value_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
(cstan_2_1_4$P_value_A_Mock_high_vs_Mock_low > 0.05 &
cstan_2_1_4$P_value_G_Mock_high_vs_Mock_low < 0.05) ~
cstan_2_1_4$Flag_G_Mock_high_vs_Mock_low,
!is.na(cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_1_4$Flag_A_Mock_high_vs_Mock_low
)
)
attr(cstan_2_1_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_1_4, "data_class") <- "pepData"
# main effects: 2; covariates: 2; groups: 4 ---------------
# Combine the G-test and ANOVA results and place columns in correct order.
cstan_2_2_4 <- dplyr::full_join(gstan_2_2_4[, c(1:5, 16:27)],
astan_2_2_4) %>%
dplyr::relocate(
dplyr::starts_with("Count_", vars = colnames(.data)),
.after = "Mass_Tag_ID"
) %>%
dplyr::relocate(
dplyr::starts_with("Mean_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Fold_change_", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("P_value_G", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_A", vars = colnames(.data)),
.after = dplyr::last_col()
) %>%
dplyr::relocate(
dplyr::starts_with("Flag_G", vars = colnames(.data)),
.after = dplyr::last_col()
)
# Replace all NaN values with NA.
cstan_2_2_4[is.nan(data.matrix(cstan_2_2_4))] <- NA
class(cstan_2_2_4) <- c("statRes", "data.frame")
attr(cstan_2_2_4, "group_DF") <- groupDF_2_2_4
attr(cstan_2_2_4, "comparisons") <- c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low")
attr(cstan_2_2_4, "number_significant") <- data.frame(
Comparison = c("Infection_high_vs_Infection_low",
"Infection_high_vs_Mock_high",
"Infection_high_vs_Mock_low",
"Infection_low_vs_Mock_high",
"Infection_low_vs_Mock_low",
"Mock_high_vs_Mock_low"),
Up_total = c(sum(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 1] == 1))),
sum(length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 2] == 1))),
sum(length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 3] == 1))),
sum(length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 4] == 1))),
sum(length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 5] == 1))),
sum(length(which(flag_g_2_2_4[, 6] == 1)),
length(which(flag_a_2_2_4[, 6] == 1)))),
Down_total = c(sum(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 1] == -1))),
sum(length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 2] == -1))),
sum(length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 3] == -1))),
sum(length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 4] == -1))),
sum(length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 5] == -1))),
sum(length(which(flag_g_2_2_4[, 6] == -1)),
length(which(flag_a_2_2_4[, 6] == -1)))),
Up_anova = c(length(which(flag_a_2_2_4[, 1] == 1)),
length(which(flag_a_2_2_4[, 2] == 1)),
length(which(flag_a_2_2_4[, 3] == 1)),
length(which(flag_a_2_2_4[, 4] == 1)),
length(which(flag_a_2_2_4[, 5] == 1)),
length(which(flag_a_2_2_4[, 6] == 1))),
Down_anova = c(length(which(flag_a_2_2_4[, 1] == -1)),
length(which(flag_a_2_2_4[, 2] == -1)),
length(which(flag_a_2_2_4[, 3] == -1)),
length(which(flag_a_2_2_4[, 4] == -1)),
length(which(flag_a_2_2_4[, 5] == -1)),
length(which(flag_a_2_2_4[, 6] == -1))),
Up_gtest = c(length(which(flag_g_2_2_4[, 1] == 1)),
length(which(flag_g_2_2_4[, 2] == 1)),
length(which(flag_g_2_2_4[, 3] == 1)),
length(which(flag_g_2_2_4[, 4] == 1)),
length(which(flag_g_2_2_4[, 5] == 1)),
length(which(flag_g_2_2_4[, 6] == 1))),
Down_gtest = c(length(which(flag_g_2_2_4[, 1] == -1)),
length(which(flag_g_2_2_4[, 2] == -1)),
length(which(flag_g_2_2_4[, 3] == -1)),
length(which(flag_g_2_2_4[, 4] == -1)),
length(which(flag_g_2_2_4[, 5] == -1)),
length(which(flag_g_2_2_4[, 6] == -1))),
row.names = NULL
)
attr(cstan_2_2_4, "statistical_test") <- "combined"
attr(cstan_2_2_4, "adjustment_method_a_multcomp") <- "none"
attr(cstan_2_2_4, "adjustment_method_g_multcomp") <- "none"
attr(cstan_2_2_4, "adjustment_method_a_fdr") <- "none"
attr(cstan_2_2_4, "adjustment_method_g_fdr") <- "none"
attr(cstan_2_2_4, "pval_thresh") <- 0.05
attr(cstan_2_2_4, "data_info") <- list(
data_scale_orig = "abundance",
data_scale = "log",
norm_info = list(is_normalized = FALSE),
num_edata = length(unique(gfilta_2_2_4$e_data$Mass_Tag_ID)),
num_miss_obs = sum(is.na(gfilta_2_2_4$e_data[, -1])),
prop_missing = (sum(is.na(gfilta_2_2_4$e_data[, -1])) /
prod(dim(gfilta_2_2_4$e_data[, -1]))),
num_samps = dim(gfilta_2_2_4$f_data)[1],
data_types = NULL
)
attr(cstan_2_2_4, "which_X") <-
c(0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0,
1, 0, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, NA, 0, 0, NA, 0, 0, 1,
NA, 0, 0, 0, 1, 0, 0, 0, 0, 0, NA, NA, 0, 0, 0, 0, 0, NA, NA,
NA, 0, 0, 0, NA, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)
attr(cstan_2_2_4, "bpFlags") <- data.frame(
Mass_Tag_ID = gfilta_2_2_4$e_data$Mass_Tag_ID,
Infection_high_vs_Infection_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
(cstan_2_2_4$P_value_A_Infection_high_vs_Infection_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Infection_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Infection_low,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Infection_low
),
Infection_high_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_high > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Mock_high < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_high,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Mock_high
),
Infection_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
(cstan_2_2_4$P_value_A_Infection_high_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_high_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_high_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Infection_high_vs_Mock_low
),
Infection_low_vs_Mock_high = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
is.na(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_high > 0.05 &
cstan_2_2_4$P_value_G_Infection_low_vs_Mock_high < 0.05) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_high,
!is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high) ~
cstan_2_2_4$Flag_A_Infection_low_vs_Mock_high
),
Infection_low_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
(cstan_2_2_4$P_value_A_Infection_low_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Infection_low_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Infection_low_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Infection_low_vs_Mock_low
),
Mock_high_vs_Mock_low = dplyr::case_when(
is.na(cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
is.na(cstan_2_2_4$P_value_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
(cstan_2_2_4$P_value_A_Mock_high_vs_Mock_low > 0.05 &
cstan_2_2_4$P_value_G_Mock_high_vs_Mock_low < 0.05) ~
cstan_2_2_4$Flag_G_Mock_high_vs_Mock_low,
!is.na(cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low) ~
cstan_2_2_4$Flag_A_Mock_high_vs_Mock_low
)
)
attr(cstan_2_2_4, "cnames") <- list(
edata_cname = "Mass_Tag_ID",
emeta_cname = NULL,
fdata_cname = "SampleID",
techrep_cname = NULL
)
attr(cstan_2_2_4, "data_class") <- "pepData"
# Save standards for IMD-ANOVA tests -------------------------------------------
# save() the following objects in "inst/testdata/standards_imd_anova.RData"
# astan_1_0_2,
# gstan_1_0_2,
# cstan_1_0_2,
# astan_1_1_3,
# gstan_1_1_3,
# cstan_1_1_3,
# astan_1_2_3,
# gstan_1_2_3,
# cstan_1_2_3,
# astan_2_0_3,
# gstan_2_0_3,
# cstan_2_0_3,
# astan_2_1_4,
# gstan_2_1_4,
# cstan_2_1_4,
# astan_2_2_4,
# gstan_2_2_4,
# cstan_2_2_4,
# tukey_pval_1_1_3,
# tukey_pval_1_2_3,
# tukey_pval_2_0_3,
# tukey_pval_2_1_4,
# tukey_pval_2_2_4,
# dunnett_1_1_3,
# dunnett_1_2_3,
# dunnett_2_0_3,
# dunnett_2_1_4,
# dunnett_2_2_4
# DO NOT PUT CODE HERE TO SAVE THE ABOVE OBJECTS. It may create problems with CRAN submission.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.