Question

is it possible to use TPM in DESeq2 ?

0

Entering edit mode

Hicham ▴ 10

@2952f756

Last seen 15 months ago

Morocco

My question is quite simple, I want to know is it possible to use TPM data to normalize them with DESeq2? and why?

counts Normalization DESeq2 • 8.1k views

ADD COMMENT • link updated 23 months ago by swbarnes2 ★ 1.3k • written 23 months ago by Hicham ▴ 10

score 1 · Answer 1 · 2022-05-10

1

Entering edit mode

Kevin Blighe ★ 3.9k

@kevin

Last seen 7 days ago

Republic of Ireland

No. DESeq2 requires raw count data as input.

If you would like to use TPM, then please take a look at these previous answers, which additionally link to other answers:

Kevin

ADD COMMENT • link 23 months ago Kevin Blighe ★ 3.9k

0

Entering edit mode

I found that data I downloaded is quantified using RSEM is it also not directly useful for DESeq2 ?

the description related to my data "tab-delimited data matrix TPM quantified by RSEM for all samples"

ADD REPLY • link 23 months ago Hicham ▴ 10

0

Entering edit mode

Expected counts from RSEM is fine for DESeq. With tximport, you can put the RSEM input into DESeq, and it will even apply offsets for gene length.

ADD REPLY • link 23 months ago swbarnes2 ★ 1.3k

0

Entering edit mode

Ah ha, yes, coming via the tximport route and RSEM is okay.

ADD REPLY • link 23 months ago Kevin Blighe ★ 3.9k

0

Entering edit mode

But it sounds like the OP only has the TPM pulled from RSEM, and not the whole output.

ADD REPLY • link 23 months ago swbarnes2 ★ 1.3k