Question

Fold Change for rnaseq DTU

0

Entering edit mode

rbenel ▴ 40

@rbenel-13642

Last seen 2.3 years ago

Israel

Hi!

I have been using the rnaseqDTU workflow and the function testForDEU internally uses the DESeq2 function nbinormLRT. I was wondering what the ramifications are if I wanted to calculate fold change for example for specific timepoints.. Hypothetically could I use the wald test instead of LRT in this case to get FCs?

Thanks!

dxd <- estimateSizeFactors(dxd) 
  dxd <- estimateDispersions(dxd, quiet=FALSE)
  dxd <- testForDEU(dxd, reducedModel=~sample + exon + replicate:exon)

DEXSeq rnaseqDTU • 1.6k views

ADD COMMENT • link 3.3 years ago rbenel ▴ 40

score 1 · Answer 1 · 2021-08-23

1

Entering edit mode

Michael Love 43k

@mikelove

Last seen 27 minutes ago

United States

I’m not sure DEXSeq offers access to those individual LFCs but I’ll defer to the package authors.

ADD COMMENT • link 3.3 years ago Michael Love 43k

0

Entering edit mode

OK. Thanks Michael! Should I add an additional tag to this post?

ADD REPLY • link 3.3 years ago rbenel ▴ 40

0

Entering edit mode

Any update on this post authors of DEXSeq?

ADD REPLY • link 3.3 years ago rbenel ▴ 40

1

Entering edit mode

I pinged the authors.

ADD REPLY • link 3.3 years ago Michael Love 43k

0

Entering edit mode

Thank You!

ADD REPLY • link 3.3 years ago rbenel ▴ 40

0

Entering edit mode

This is a good point. It should be possible to use the Wald test vs LRT to get the LFC of the interaction exon:condition. But I've not compared the different approaches myself. If any (differences) I guess they would be similar to those observed for gene level tests.