RunVoomLimma

RunVoomLimma

Usage

RunVoomLimma(dgelists, design_matrix, do_contrast_fit, my_contrast_matrix)

Arguments

dgelists

a ist of DGElist created in NormalizePseudobulk

design_matrix

design matrix

do_contrast_fit

whether to fit custom a priori contrasts

my_contrast_matrix

custom a priori contrasts created with make.contrasts - see limma or edgeR manual

Value

a list of linear model fits for each celltype

Examples

if (FALSE) {
c_mat = makeContrasts(
foldchange_difference = (group_timepoint1_1 - group_timepoint1_0) - (group_timepoint0_1 - group_timepoint0_0),
time1_foldchange = (group_timepoint1_1 + group_timepoint0_1) / 2  - (group_timepoint1_0 + group_timepoint0_0) / 2,
baseline_groups = (group_timepoint1_0 - group_timepoint0_0),
levels = colnames(designmat)
)
# fit simple linear model for the baseline group level contrast
bl = scglmmr::RunVoomLimma(dgelists = dge,
design_matrix = designmat,
do_contrast_fit = T,
my_contrast_matrix = c_mat[ ,3])
}
# run limma using voom observational weights for non mixed effects models using emperical bayes