Hello everyone, I ran nf-core/rnaseq pipeline successfully, now I want to do downstream analysis like machine learning and so on. Should I use salmon.merged.gene_counts.tsv for that and applying log transformation to normalize it before doing downstream analysis?
There are multiple count tables from nf-core/rnaseq, the best depends on your use case. I would personally use the Salmon pseudoalignment counts and import them with tximport into deseq2 or limma-voom. I wouldn’t transform the data before doing analysis, see discussions like this.
For a downstream analysis tool, try nf-core/differentialabundance
1 Like