EdgeR

edgeR Package

Differential expression analysis for RNA-seq.

Basic Workflow

library(edgeR)

# Create DGEList
dge <- DGEList(counts = count_matrix, group = group)

# Filter low counts
keep <- filterByExpr(dge)
dge <- dge[keep, , keep.lib.sizes = FALSE]

# Normalize
dge <- calcNormFactors(dge)

# Estimate dispersion
dge <- estimateDisp(dge)

# Exact test (two groups)
et <- exactTest(dge)
topTags(et)

GLM Approach

# Design matrix
design <- model.matrix(~0 + group)
colnames(design) <- levels(group)

# Estimate dispersion
dge <- estimateDisp(dge, design)

# Fit GLM
fit <- glmQLFit(dge, design)

# Contrast
contrast <- makeContrasts(TreatmentA - Control, levels = design)
qlf <- glmQLFTest(fit, contrast = contrast)
topTags(qlf)

Results

# Top genes
results <- topTags(qlf, n = Inf)$table

# Significant genes
sig <- results[results$FDR < 0.05, ]

# Volcano plot
plotMD(qlf)

# MA plot
plotSmear(qlf)

Normalization

# TMM (default)
dge <- calcNormFactors(dge, method = "TMM")

# Other methods
dge <- calcNormFactors(dge, method = "RLE")
dge <- calcNormFactors(dge, method = "upperquartile")

# Get normalized counts
cpm(dge)
rpkm(dge, gene.length = gene_lengths)

MDS Plot

plotMDS(dge, col = as.numeric(group))

Export

write.csv(topTags(qlf, n = Inf)$table, "de_results.csv")

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: LeoLin990405
• Source: LeoLin990405/r-analytics-skill
• License: MIT

Install and usage instructions live in the source repository linked above.