search

zhangyuqing / ComBat-seq

Batch effect adjustment based on negative binomial regression for RNA sequencing count data
154 stars 39 forks source link

including more than one feature to be corrected #8

Closed leonorrib closed 3 years ago

leonorrib commented 4 years ago

Hello, thank you for developing ComBat-seq. It is definitely useful. I would like to ask whether there is a way to use ComBat-seq to correct two different features. One feature would be the typical batch effect, and another would be to correct for different sample concentrations in the libraries.

Best, Leonor

zhangyuqing commented 3 years ago

Hi @leonorrib, thanks for your comment, interesting question! ComBat-Seq is not designed to resolve sample concentration differences by itself. However, there may be ways to address it, for example, cluster samples based on sample concentrations, then treat the clusters as a type of "batch". This would require future research.