permutationMC: Permute group labels

In aametwally/OmicsLonDA: Omics Longitudinal Differential Analysis

Description

Permutes the group label of the samples in order to construct the testStatistics empirical distibution

Usage

permutationMC(formula = Count ~ Time, perm.dat = NULL, n.perm = 500,
  fit.method = "ssgaussian", points, parall = FALSE, prefix = "Test")

Arguments

formula	formula to be passed to the regression model
perm.dat	dataframe has the Count, Group, Subject, Time
n.perm	number of permutations
fit.method	The fitting method (ssgaussian)
points	The points at which the prediction should happen
parall	boolean to indicate whether to use multicore.
prefix	prefix to be used to create directory for the analysis results

Value

a list of the fitted model for each group for all the permutations

References

Ahmed Metwally ([e-mail])

Examples

library(SummarizedExperiment)
data("omicslonda_data_example")
omicslonda_se_object_adjusted = adjustBaseline(
    se_object = omicslonda_data_example$omicslonda_se_object)
omicslonda_test_object = omicslonda_se_object_adjusted[1,]
se_object = omicslonda_test_object
dt = data.frame(colData(se_object))
dt$Count = as.vector(assay(se_object))
Group = as.character(dt$Group)
group.levels = sort(unique(Group))
gr.1 = as.character(group.levels[1])
gr.2 = as.character(group.levels[2])
df = dt
levels(df$Group) = c(levels(df$Group), "0", "1")
df$Group[which(df$Group == gr.1)] = 0
df$Group[which(df$Group == gr.2)] = 1
group.0 = df[df$Group == 0, ]
group.1 = df[df$Group == 1, ]
points = seq(100, 130)
perm  = permutationMC(formula = Count ~ Time, perm.dat = df, n.perm = 10,
                      fit.method = "ssgaussian", points = points,
                      parall = FALSE, prefix = tempfile())

aametwally/OmicsLonDA documentation built on Feb. 10, 2020, 11:47 a.m.