search

zhangyuqing / ComBat-seq

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

ComBat-Seq and pseudobulk counts #12

Closed AmelZulji closed 4 years ago

AmelZulji commented 4 years ago

Thank you for this major improvement, Yuqing!

Since you mentioned in the previous issues that it is valid to use corrected values as input for DESeq2, I am wondering can the ComBat-seq be used for correcting pseudobulk-counts from single-cell data followed by DESeq2 DGE analysis?

In addition, if experimental design includes control and condition samples processed in multiple batches, is separate batch correction (i.e. separate batch correction of control samples + separate batch correction of condition samples, followed by merging matrices) a valid approach? Or that can be handled by specifying groups?

Thank you in advance for your help!

Kind regards, Amel Zulji

zhangyuqing commented 4 years ago

Hi @AmelZulji, yes, you can use ComBat-seq to correct pseudobulk-counts. The control and condition samples can be handled by specifying groups in the parameters.

AmelZulji commented 4 years ago

Thank you very much, Yuqing!