nhuhoa
commented
2 years ago Hi,
RNA-seq means bulk RNA-seq? I think the best and also classical methods for bulk RNA-seq are ComBat or limma. I see another method that was particularly designed for bulk is Combat-seq but I never tried it before. You may want to test it.
https://github.com/zhangyuqing/ComBat-seq
From my point of view, bulk data is different from single cell data, and a good normalization method that allow to remove library size effects (sequencing depth, total read counts) is good enough to go. Ex: using edgeR: https://www.bioconductor.org/packages/release/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf
- combine all bulk samples into 1 data frame
- filtering genes with min_counts over all samples, filterbyExpr()
- compute library size
- calcNormFactors
- then convert normalized data to CPM table
- you may want to see more at: RNA-Seq of pathogen inoculated arabidopsis with batch effects https://www.bioconductor.org/packages/release/bioc/vignettes/edgeR/inst/doc/edgeRUsersGuide.pdf in case you have many bulk samples
Have fun, Hoa
What do you recommend to try out for "big data" but for RNA seq instead of single cell?