How to use the Combat-seq for batch effect removal?

Question

[Yoomi]

I try to remove the batch effect using Combat-seq in RNA-seq count data. Should I normalize RNA-seq raw count data before or after using the Combat-seq? Show me the example code of how to use Combat-seq and Normalization.

[Ram]

Show me the example code of how to use Combat-seq and Normalization.

This might be a language barrier, but please do not use the imperative tone without any "please" or "could you" or such polite phrases. It comes across as demanding and on forums, could lead to antagonizing people unintentionally.

[Yoomi]

Thank you so much~!!:)

Answer 1

[ATpoint]

Combat-seq accepts raw counts and returns batch-corrected raw counts. Once you have the corrected counts treat them as any raw counts with any RNA-seq package of choice.

[Yoomi]

Thank you so much~!^^

Answer 2

[bk11]

should i normalized the RNA-seq raw countdata before or after using the Combat-seq?

Please see in this post below which had been discussed earlier.

In Which Order Use Normalization And Batch Effects Removal?

Also remember that standard methods like DESeq2, edgeR etc. can handle batch effect very well. You do not have to perform batch correction using outside softwares if you are planning to perform differential expression analysis using these methods.

[Yoomi]

Thank you~!:)

Can I handle the batch effect in this manner?

dds.dex <- DESeqDataSetFromMatrix(countData = round(ga.ta.dex.merge), colData = coldata, design = ~ condition)
dds.dex <- DESeq(dds.dex)
vsd <- vst(dds.dex, blind=FALSE)
mat <- assay(vsd)
mm <- model.matrix(~condition, colData(vsd))
mat <- limma::removeBatchEffect(mat, batch=vsd$batch , design=mm)
assay(vsd) <- mat