Using Limma in R to get log2 fold change data for GSE series data, want to get rid of iso-form expression value

Question

[sslimb95]

Hi, every one.

I'm a beginner in DEG.

I want to get a DEG data which have log2 fold change expression values or GSE29221 using Limma in R. And finally, i got a DEG data, but there was a problem.

In DEG data, there are many isoforms' expression values. For example, a gene which entrez gene id is '1' have two log2 fold change values which are -2.9 and -1.4.

Is there any solution that i have only one log2 fold change value for each genes? I don't want to have other isoforms' expression values.

Many thanks.

Answer 1

[Alex Nesmelov]

You can recount transcription for genes from bam files, using htseq2 with '-t gene' option. For the calculated expression data, limma::avereps can average data for transcripts (https://www.biostars.org/p/415292/). I'd prefer the first option. By the way, is '1' really an Entrez gene id?

[sslimb95]

Yes it is. https://www.genecards.org/cgi-bin/carddisp.pl?gene=A1BG

AIBG gene's external id which is entrez gene id is 1.

And the expression data were

Entrez gene ID logFC AveExpr t P.Value adj.P.Val B

1 -2.906807042 1.235062938 -1.627652551 0.117324755 0.433154895 -4.38636393

1 -1.373079717 4.84894835 -0.570154359 0.574142708 0.848593976 -5.306552283