Hi 

I am new user to DESeq2.  I managed to perform the analysis.  But i want to build heatmaps based on the pathway changes.  Say for example i can extract ids which are significantly changed lfc > 2 and padj < 0.05.  How can i use those IDs to extract counts from the DESeqDataSet and plot them as heatmap?

For example using the vignette, 

library("genefilter")

library(pheatmap)

topVarGenes <- head(order(-rowVars(assay(rld))),20)

mat <- assay(rld)[ topVarGenes, ]
mat <- mat - rowMeans(mat)
df <- as.data.frame(colData(rld)[,c("Condition", "libType")])

pheatmap(mat, fontsize_number = 0.8 * fontsize, annotation_col=df)

Using this script i can easily draw a heatmap of top 50 genes.  But for example if i have a list of genes of interest and i if i want to draw a heatmap by would it be possible to do so.

thanks in advance

nishikant 

hi Nishikant,

assay(rld) has rownames which identify the genes. To make it clearer, let's assign this to a new variable, so it's clear it's just a matrix:

mat <- assay(rld)

Now look at the first few rownames:

head(rownames(mat))

You can extract the rows you want by indexing, you just need to create a vector of the genes you want to pull out. For example:

idx <- rownames(res)[ which(res$padj < 0.1) ]

Then this is how you would index mat:

mat[ idx, ]