Why 2000 out of 7000 genes are not calculated?

Question

[Parham]

Hi,

I have S. pombe as organism and I have RNA-seq data from it mapped and checked for gene enrichment analysis using this workflow. However there are two things that concerns me. First when I remove rows with zero sum, total number of genes that I have drops from 7018 to about 6000 genes.

> cnts=hnrnp.cnts
> sel.rn=rowSums(cnts) != 0
> cnts=cnts[sel.rn,]

Next thing when I use joint workflow DESeq2 and I finally create the deseqres there are about another 1000 genes with NAin padjcolumn. So I guess I can't use those genes in my analysis further down the workflow. In total I am missing about 2000 out of 7000 of my genes which seems a lot. I wanted to know if there is something wrong with my analysis or anything that experts are aware of.

I am not bioinformatician!

Thanks!

Answer 1

[Michael Love]

Hi,

Read over the help file

?results 

and the vignette:

browseVignettes("DESeq2")

It is also discussed in the manuscript under "independent filtering".