1
$\begingroup$

Is it possible to use sva's ComBat for batch effects removal on scRNA-seq data?

Is there any difference between RNA-seq and scRNA-seq data which doesn't allow to use ComBat on single-cell data? (I am thinking about extreme sparsity, > 80%). If so, are there alternatives?

$\endgroup$
2
  • 1
    $\begingroup$ see here for a simiar [unanswered] question. $\endgroup$ – gringer Feb 2 '18 at 19:43
  • $\begingroup$ @gringer Indeed, that was another question of mine :) $\endgroup$ – gc5 Feb 2 '18 at 20:54
5
$\begingroup$

Look at this recent paper that uses ComBat on scRNA-seq data for batch effect removal and states that it "successfully does so".

I also suggest that you check out this publication on Distribution Matching Residual-Nets. Authors evaluated their method also on scRNA-seq data and thus it may be something you are looking for. I personally played a bit with their code and found it straightforward to use.

$\endgroup$
4
$\begingroup$

There are competing claims regarding how well ComBat works on scRNA-seq datasets. A recent paper from John Marioni's introduces a mutual nearest neighbors method that seems to outperform ComBat in at least some relevant scenarios. In general, it's best to look at tSNE or other diagnostic plots after batch correction to see if the results seem reasonable (this advice holds generally).

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.