Dear all,
I'd like to combine different datasets (TCGA-GBM, TCGA-LGG, SRP027383, GTEX brain) from recount-brain to perform differential gene expression analysis with DESeq2 (+ use normalized counts for visualization purposes).
So far I've studied following topics and workflows:
https://support.bioconductor.org/p/9151174/
https://support.bioconductor.org/p/9143498/
http://research.libd.org/recount-brain/example_multistudy/recount_brain_multistudy.html
http://research.libd.org/recount-brain/example_SRP027383/example_SRP027383.html
But I'm still not sure whether it's possible and if yes, how to do it. Do I need to separate glioblastoma from control tissues prior to normalizing the data to remove dataset specific effects? Separate tumors depending on grading?
How about normalizing for dataset and adjusting for PCs before feeding the data to DESeq2 as in: http://research.libd.org/recount-brain/example_multistudy/recount_brain_multistudy.html#normalize-for-dataset (Points 3.4 and 3.5)
Contrary to this topic: https://support.bioconductor.org/p/9151174/ some authors seemed to go through with it: https://www.nature.com/articles/s41598-023-31180-z
Would running ComBat-seq with passing dataset to covar_mod be sufficient prior to DESeq analysis? (after verifying with PCA)
I'd really apreciate your guidance with this!
