Hello Bioinformatics Community,

I'm currently analyzing an RNA-seq dataset involving subtypes of disease from 16 brain tissue samples, with 2 runs each making 32 SRR runs. Each biological sample has multiple sequencing runs, one sample has two runs, resulting in technical replicates. I'm seeking guidance on the optimal strategy to incorporate these replicates into my differential expression analysis.

Specific Questions:

Merging Technical Replicates:Should technical replicates (multiple sequencing runs from the same biological sample) be merged:

before alignment,

after alignment but before counting, or

after obtaining gene expression counts?

By merging, I mean should I add gene counts?

Downstream Analysis (DESeq2/edgeR):What is the recommended method for handling these technical replicates to ensure accurate and robust differential expression results? Should I use functions such as collapseReplicates (DESeq2) or sumTechReps (edgeR)?

Any recommendations, protocols, or references would be greatly appreciated.

Thank you!

Please don't cross-post. You have already asked this on biostars.org, which is the correct venue for general questions like this.