fitNBthDE-methods: Negative Binomial threshold model for differential expression...

In Nanostring-Biostats/GeoDiff: Count model based differential expression and normalization on GeoMx RNA data

Negative Binomial threshold model for differential expression analysis

Description

Negative Binomial threshold model for differential expression analysis

Negative Binomial threshold model for differential expression analysis

Usage

fitNBthDE(object, ...)

## S4 method for signature 'NanoStringGeoMxSet'
fitNBthDE(
  object,
  form,
  split,
  ROIs_high = NULL,
  features_high = NULL,
  features_all = NULL,
  sizefact_start = NULL,
  sizefact_BG = NULL,
  threshold_mean = NULL,
  preci2 = 10000,
  lower_threshold = 0.01,
  prior_type = c("contrast", "equal"),
  sizefactrec = TRUE,
  size_scale = c("sum", "first"),
  sizescalebythreshold = FALSE,
  iterations = 2,
  covrob = FALSE,
  preci1con = 1/25,
  cutoff = 10,
  confac = 1
)

## S4 method for signature 'matrix'
fitNBthDE(
  form,
  annot,
  object,
  probenum,
  features_high,
  features_all,
  sizefact_start,
  sizefact_BG,
  threshold_mean,
  preci2 = 10000,
  lower_threshold = 0.01,
  prior_type = c("contrast", "equal"),
  sizefactrec = TRUE,
  size_scale = c("sum", "first"),
  sizescalebythreshold = FALSE,
  iterations = 2,
  covrob = FALSE,
  preci1con = 1/25,
  cutoff = 10,
  confac = 1
)

Arguments

object	count matrix with features in rows and samples in columns
...	additional argument list that might be used
form	model formula
split	indicator variable on whether it is for multiple slides (Yes, TRUE; No, FALSE)
ROIs_high	ROIs with high expressions defined based on featfact and featfact
features_high	subset of features which are well above the background
features_all	full list of features
sizefact_start	initial value for size factors
sizefact_BG	size factor for background
threshold_mean	average threshold level
preci2	precision for the background, default=10000
lower_threshold	lower limit for the threshold, default=0.01
prior_type	empirical bayes prior type, choose from c("contrast", "equal")
sizefactrec	whether to recalculate sizefact, default=TRUE
size_scale	method to scale the sizefact, sum(sizefact)=1 when size_scale="sum", sizefact[1]=1 when size_scale="first"
sizescalebythreshold	XXXX, default = FALSE
iterations	how many iterations need to run to get final results, default=2, the first iteration apply the model only on features_high and construct the prior then refit the model using this prior for all genes.
covrob	whether to use robust covariance in calculating covariance. default=FALSE
preci1con	The user input constant term in specifying precision matrix 1, default=1/25
cutoff	term in calculating precision matrix 1, default=10
confac	The user input factor for contrast in precision matrix 1, default=1
annot	annotations files with variables in the formula
probenum	a vector of numbers of probes in each gene, default = rep(1, NROW(object))

Value

a list of

• X, design matrix

• para0, estimated parameters for the first iteration, including regression coefficients, r and threshold in rows and features in columns

• para, estimated parameters, including regression coefficients, r and threshold in rows and features in columns

• sizefact, estimated sizefact

• sizefact0, estimated sizefact in iter=1

• preci1, precision matrix for regression coefficients estimated in iter=1

• Im0, Information matrix of parameters in iter=1

• Im, Information matrix of parameters in iter=2

• conv0, vector of convergence for iter=1, 0 converged, 1 not converged

• conv, vector of convergence for iter=2, 0 converged, 1 not converged

• features_high, same as the input features_high

• features_all, same as the input features_all

a list of

• X, design matrix

• para0, estimated parameters for the first iteration, including regression coefficients, r and threshold in rows and features in columns

• para, estimated parameters, including regression coefficients, r and threshold in rows and features in columns

• sizefact, estimated sizefact

• sizefact0, estimated sizefact in iter=1

• preci1, precision matrix for regression coefficients estimated in iter=1

• Im0, Information matrix of parameters in iter=1

• Im, Information matrix of parameters in iter=2

• conv0, vector of convergence for iter=1, 0 converged, 1 not converged

• conv, vector of convergence for iter=2, 0 converged, 1 not converged

• features_high, same as the input features_high

• features_all, same as the input features_all

Examples

library(Biobase)
library(dplyr)
data(demoData)
demoData <- demoData[, c(1:5, 33:37)]
demoData <- fitPoisBG(demoData, size_scale = "sum")
demoData <- aggreprobe(demoData, use = "cor")
demoData <- BGScoreTest(demoData)
demoData$slidename <- substr(demoData[["slide name"]], 12, 17)
thmean <- 1 * mean(fData(demoData)$featfact, na.rm = TRUE)
demo_pos <- demoData[which(!fData(demoData)$CodeClass == "Negative"), ]
demo_neg <- demoData[which(fData(demoData)$CodeClass == "Negative"), ]
sc1_scores <- fData(demo_pos)[, "scores"]
names(sc1_scores) <- fData(demo_pos)[, "TargetName"]
features_high <- ((sc1_scores > quantile(sc1_scores, probs = 0.4)) &
   (sc1_scores < quantile(sc1_scores, probs = 0.95))) |>
    which() |>
    names()
set.seed(123)
demoData <- fitNBth(demoData,
                    features_high = features_high,
                    sizefact_BG = demo_neg$sizefact,
                    threshold_start = thmean,
                    iterations = 5,
                    start_para = c(200, 1),
                    lower_sizefact = 0,
                    lower_threshold = 100,
                    tol = 1e-8)
ROIs_high <- sampleNames(demoData)[which(demoData$sizefact_fitNBth * thmean > 2)]
features_all <- rownames(demo_pos)

pData(demoData)$group <- c(rep(1, 5), rep(2, 5))

NBthDEmod1 <- fitNBthDE(
    form = ~group,
    split = FALSE,
    object = demoData,
    ROIs_high = ROIs_high,
    features_high = features_high,
    features_all = features_all,
    sizefact_start = demoData[, ROIs_high][["sizefact_fitNBth"]],
    sizefact_BG = demoData[, ROIs_high][["sizefact"]],
    preci2 = 10000,
    prior_type = "contrast",
    covrob = FALSE,
    preci1con = 1/25,
    sizescalebythreshold = TRUE
)

Nanostring-Biostats/GeoDiff documentation built on April 24, 2022, 5:54 p.m.