2.10 Differential expression analysis
Estimated read counts from featureCounts were used as input to functions
in the DESeq2 R package (Love et al. 2014) to generate log2 differential
expression fold-difference estimates. Transcripts with less than 10
reads summed across all samples were removed from the analysis. Genes
with adjusted p-value less than 5% (according to the FDR method from
Benjamini-Hochberg, 1995) were declared differentially expressed. A
log2FoldChange value greater than zero indicates an upregulation in FW
compared to SW and a log2FoldChange value less than 0 indicates a
downregulation in FW compared to SW. In order to explore the variability
within our experiment, hierarchical clustering and PCA were performed
after Variance Stabilizing Transformation (VST) of the count data.