batch effect
0
0
Entering edit mode
Syed Zaheer ▴ 10
@ec7085d4
Last seen 6 days ago
Pakistan

Hi deseq2 is good considering batches and I can see and remove them with a command assay(vsd) <- limma::removeBatchEffect(assay(vsd), vsd\$Batch) and see with plot but how can I see the deseq2 summary considering this removal I think deseq2 is not taken into account this thing. how can i do this please reply thanks in advance



TargetScoreData batchelor • 142 views
0
Entering edit mode

Please edit the question, try to make clear what the question is. Code and example data are appreciated. If the question is how you can check whether DESeq2 is respecting a design of e.g. ~batch+group (so whether batch is respected) then be assured it is. The package is in Bioc almost 10 years, widely used and well-tested. There is no direct way of checking that, this is why the vignette recommends to use the removeBatchEffect` strategy to get a "visualization" of how regressing batches (or any covariate) behaves. If you have a design with a batch in it and DESeq2 runs without complaining then it is taken into account. What you can of course do is to run the analysis with and without batch and simply compare the results tables. They will be different, maybe (very) slightly if the batch effect is minimal, but they will be different.

0
Entering edit mode

Thanks for the reply with a comprehensive explanation. yes, I did run with and without and got different results.

thank you... but the last thing I am new to these analyses. would be silly question its not related to deseq2 question is I am expecting fewer differential expressed genes with batches. but with batch term no of genes differential expressed increased

and I put more number of Control samples number of genes differentially expressed are increasing aah what ...

0
Entering edit mode

It is not unusual that you get more differential genes if you account for batch effects. Think about it, usually a batch effect is an unwanted variation in the data and that drives up the dispersion (makes sample difference even within the same group larger than it actually is)...and that often leads to fewer DE genes. hence correcting for batch often helps getting more DE genes.

1
Entering edit mode

excellent Thanks a lot everything worked perfectly :-)