DESeQ2 design multiple conditions
1
0
Entering edit mode
@domenicoalessandrosilvestris-20812
Last seen 4.8 years ago

Hi,

I'm using DESeq2 with 4 multiple conditions ("A", "B", "C" and "D"), but I'm only interested in the comparisons A vs C; A vs D; B vs C and B vs D. Since the higher the number of tests performed and the lower the significant results (and the "contrast" argument of the result function performs only a filtering I think) I wondered if there was a way to set the "desing" avoiding that are all possible comparisons in pairs between conditions performed? Thank you very much in advance to anyone who wants to help me !!

Alessandro

deseq2 • 740 views
ADD COMMENT
0
Entering edit mode
swbarnes2 ★ 1.3k
@swbarnes2-14086
Last seen 8 hours ago
San Diego

Have you read any vignettes or tutorials? If you are thinking that the default setting is to do all 6 comparisons and mash the results together, it's not. The simplest way to run the software is to specify which two groups you want to compare to each other, and do that 4 times to get 4 sets of results.

ADD COMMENT
0
Entering edit mode

Thanks for the reply! I know that your solution is obviously the simplest but in that way when you extract the normalized counts you get slightly different values ​​for the same gene in the same sample depending on the comparison you are considering and if you have to use these values ​​for example then in a correlation you will have a little problem..

Alessandro

0
Entering edit mode

No, you don't. You don't have to subset the data to compare one sub group to another. Run DESeq on the whole dds object, then do the comparison you want with contrasts.

ADD REPLY
0
Entering edit mode

From what I understand "contrast" extracts only the results related to the comparison you are requesting, but the actual statistical test is performed by the "DESeq" function in fact when I go to open the dds it tells me "assays (6)". Actually the problems are two: to avoid making thousands of useless statistical tests (if not harmful) and to have a unique expression value for each gene in each sample. Thanks again!

Login before adding your answer.

Traffic: 701 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6