Differential gene expression overview

RNA-Seq with Bioconductor in R

Mary Piper

Bioinformatics Consultant and Trainer

heatmap fold changes

de variation

JCI paper

Introduction to dataset: Smoc2

kidney fibrosis

RNA-Seq experiment

RNA-Seq count distribution

ggplot(raw_counts) +
  geom_histogram(aes(x = wt_normal1), stat = "bin", bins = 200) +
  xlab("Raw expression counts") +
  ylab("Number of genes")

Count Distribution

Preparation for differential expression analysis: DESeq2 object

dds <- DESeqDataSetFromMatrix(countData = rawcounts,
                        colData = metadata,
                        design = ~ condition)

counts data

Preparation for differential expression analysis: metadata

# Create vectors containing metadata for the samples
genotype <- c("wt", "wt", "wt", "wt", "wt", "wt", "wt") 
condition <- c("normal", "fibrosis", "normal", 
               "fibrosis", "normal", "fibrosis", "fibrosis") 

# Combine vectors into a data frame
wt_metadata <- data.frame(genotype, wildtype)


# Create the row names with the associated sample names
rownames(wt_metadata) <- c("wt_normal3", "wt_fibrosis3", "wt_normal1", 
                "wt_fibrosis2", "wt_normal2", "wt_fibrosis4", "wt_fibrosis1")

Preparation for differential expression analysis: metadata

metadata

Let's practice!

RNA-Seq with Bioconductor in R