Hi,

I did deseq 2x2 model. When I do a pheatmap for highly variable genes (using rld) the outcome are the genes that they are not even among the differentially expressed genes.

Here is my command for pheatmap:

topVarGenes <- head(order(rowVars(assay(rld)),decreasing=TRUE),20)
mat <- assay(rld)[ topVarGenes, ]
mat <- mat - rowMeans(mat)
df <- as.data.frame(colData(rld)[,c("group","condition")])
pheatmap(mat, annotation_col=df)
 

For example for the gene with highest rld over samples, the basemean in dds is 98000 but the pvalue and padj is NA.

Can anyone help me with this issue? 

Many thanks in advance,

Rahel

topVarGenes looks at the row variance of the transformed values regardless of which samples come from which group. Differential expression testing asks whether the difference across group is large relative to the within-group variance. So these are different ways of ranking genes.