Batch effect of sequencing data that were sequenced in different locations

Question

[Omics data mining]

Hello Everyone

I have Bulk-RNASeq sequencing data and processing the data in the DESeq2 package. I am concerned about the "batch" effect due to the different sequencing runs performed at different location. Samples of group1 and group2 were sequenced separably at different locations.

How to handle

    condition   Batch
Sample1 G1  1
Sample2 G1  1
Sample3 G2  2
Sample4 G2  2

dds<-DESeqDataSetFromMatrix(countData = count_f, colData = col_data, design = ~Batch+condition)

Should I include "Batch" or not while performing the differential analysis among two groups (G1 vs G2)?

I would appreciate all the suggestions.

Answer 1

[ATpoint]

It's perfectly confounded (batch is the exact same as condition), so nothing you can do about it. It's a (sorry to be blunt) poor experimental design. All you can do is to hope that the batch effect (if it exists) does not overshadow the true biological variation. Not nesting batch and condition is really the basis for any experimental design, you should try to avoid that in the future.