Filtering Gene by mean expression

Question

[synat.keam]

Dear All,

Sorry for another post. I have RNAseq data and used DESeq2 pipeline for differntially expressed genes. Now I am doing WGCNA and the author suggest to filter genes based on mean expression rather than variance. I am wondering whether anyone has ever filtered genes based on mean expression and could suggest some code chunks.

Kind Regards,

Synat,

[kashiff007]

If you need only exp values based on mean use cufflinks.

[synat.keam]

Hi Marcro,

I'm not sure whether DESeq2 filter genes based on mean expression.

Kashiff, I will try your suggestion. Thank for you both of you for taking time to respond.

Regards,

synat

[Marco Pannone]

Hey! DESeq2 does not filter genes with low read counts across samples by default, but you need to perform such filtering step, simple as this:

#after generating a dds object
keep <- rowSums(counts(dds)) >= 10
dds <- dds[keep,]

[Marco Pannone]

You mean filtering genes with very low read counts as also suggested in the DESeq2 pipeline?