Question: Random effects in DESeq2
gravatar for c.taylor
3 months ago by
c.taylor0 wrote:

Is there any extension available/planned to the DESeq2 pipeline that would allow the inclusion of random effects in the model formula?

The dataset that I am trying to analyse includes repeat samples (at different time points) from the same individuals. However there are a large number of individuals (~200) and only a few samples per individual (total number of data points ~450). Therefore, I don’t believe that it would be advisable to fit individual as a fixed effect as it would likely lead to major overfitting. A random effect would seem perfect here and I have used it successfully in similar contexts with other simpler response variables. However as far as I am aware the standard DESeq function does not allow the inclusion of random effects. Based on other packages that fit negative binomial models, it seems that methods do exist for fitting random effects, but I don't know how to translate these to the context of DESeq.

Any suggestions much appreciated.

deseq2 • 135 views
ADD COMMENTlink modified 3 months ago by Michael Love26k • written 3 months ago by c.taylor0
Answer: Random effects in DESeq2
gravatar for Michael Love
3 months ago by
Michael Love26k
United States
Michael Love26k wrote:

This has been asked before on the support site:

ADD COMMENTlink written 3 months ago by Michael Love26k

Thanks very much for your help! I had tried using the search function within the bioconductor site, but for some reason that didn't turn up the linked question.

only gives this page and an unrelated question

ADD REPLYlink modified 3 months ago • written 3 months ago by c.taylor0
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 226 users visited in the last hour