Entering edit mode
4.5 years ago
manninm
•
0
My lab is looking for a tool to use in DEQ analysis after batch correction. I am aware of the pitfalls of DEG analysis after batch correction, and that the usual recommendation is to include batch as a co variate in the linear model, but SVAPLSseq's premier paper (https://www.sciencedirect.com/science/article/pii/S0888754318303148?via%3Dihub) seems to suggest that SVAPLSseq is capable of DEQ analysis after batch correction. Has anyone used or has experience with this package? Any opinions on the paper or using the tool in DEG analysis?
Just a general comment: Why using new tools if excellent and well-maintained tools already exist? What is the downside of including batch into the design? Are you forced to use that new package which probably has a tiny user base compared to DESeq2 or edgeR?
My lab has a number of cohorts and datasets that we wish to combine. We want to assess batch correction alongside co variate linear models, mainly to convince collaborators that it isn't the best method. But you bring up several valid points that we will consider and communicate to our collaborators.