eset_lm: Nested Linear Model Analysis

In vladpetyuk/vp.misc: Miscellaneous functions used by Vlad Petyuk for proteomics data analysis

Description

A function for analysis of sigfinicance of factors (akin ANOVA) or continuous variable (regression) using nested linear models. Shuffling-based FDR estimation is available as an option.

Usage

eset_lm(eset, form.alt, form.nul, facs = NULL, norm.coef = NULL, N = NULL)

Arguments

eset	eset (or most likely eset subclass) object
form.alt	character formulation of alternative model
form.nul	character formulation of NULL model
facs	data frame with the factors in its columns. If NULL, then pData(eset) will be used.
norm.coef	vector with sample-to-sample normalization coefficients in log2 scale.
N	number of shuffles for FDR estimation. Default is NULL, that is no shuffling-based FDR estimation. Warning! Will be N-times slower. This is critical for Windows, since paralellization is not impletemted. In case of enabling shuffling-based FDR estimation, the recommended value is at least N=1000.

Value

data.frame

effect

either max contrast in case of factor or slope in case of continuous variable

F.stat

F statistic

p.value

p-value

Examples

data(srm_msnset)
head(varLabels(msnset))
out <- eset_lm(msnset, "~ subject.type", "~ 1")
head(out)
# now with shuffling
out <- eset_lm(msnset, "~ subject.type", "~ 1", N = 100)
head(out)
head(out[order(out$p.value),])

vladpetyuk/vp.misc documentation built on June 25, 2021, 6:35 a.m.