DESeq2 multifactor design, interpreting interaction names
1
0
Entering edit mode
fargo • 0
@fargo-16931
Last seen 6.3 years ago

Hello,

I want to determine if there are genes that are differentially expressed between males and females for each treatment (none, low, and high). For gender my reference level is set as female and for treatment my reference level is set as none. Samples and script below.

sample treatment gender
1 none female
2 none female
3 none female
4 none male
5 none male
6 none male
7 low female
8 low female
9 low female
10 low male
11 low male
12 low male
13 high female
14 high female
15 high female
16 high male
17 high male
18 high male
#design matrix
>dds = DESeqDataSetFromMatrix(countData=countdata, colData=coldata, design=~gender+treatment+gender:treatment)
#set reference levels
>dds$treatment=relevel(dds$treatment, ref="none")
>dds$gender=relevel(dds$gender, ref="female")
>dds = DESeq(dds)
>res=results(dds)
>resultsNames(dds)
[1] "Intercept"           "gender_male_vs_female"  "treatment_low_vs_none"  "treatment_high_vs_none"  "gendermale.treatmentlow" "gendermale.treatmenthigh"

1) So is my design set up correctly for the question I want to answer, which genes are differentially expressed between males and females under a specific treatment?
2) And does "gendermale.treatmentlow" look for DEGs between males and females for low treatments, whereas "gendermale.treatmenthigh" looks for DEGS between males and females for high treatment?
3) Why is there no "gendermale.treatmentnone"?

deseq2 interactions multifactorial design • 1.3k views
ADD COMMENT
0
Entering edit mode
@mikelove
Last seen 16 hours ago
United States

You can use an alternate design which will make it easier for you to find the contrasts. Try ~treatment + treatment:gender

ADD COMMENT
0
Entering edit mode

I agree that the alternate design is easier to understand. I guess I am still confused what exactly the interaction coeffcients are comparing against. 

After running the alternate design my resultsNames(dds) are:

[1] "Intercept"   "treatment_low_vs_none"  "treatment_high_vs_none"  "treatmentnone.gendermale" "treatmentlow.gendermale" "treatmenthigh.gendermale"

Is res=results(dds, name=treatmentlow.gendermale) then comparing low treatment males to low treatment females? If not, then what is it comparing against?

 

 

ADD REPLY
0
Entering edit mode

In the new design each of the interaction terms is the male versus female difference for that treatment group.

ADD REPLY

Login before adding your answer.

Traffic: 659 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