How should I analyze experiments with three groups with limma?

Question

[lhwh1208]

Hi all,

I'm quite new to RNAseq analysis and I'm a bit confused. I have an experiment with three groups: A, B and C, where C is the control. Each group has three biological replicates. A and B conditions are opposite of each other (high metal and low metal concentrations). Should I run all three groups together in limma and designate A-C, B-C contrasts, or run A-C and B-C separately?

Also, I ran all three groups together and checked A-C contrast results. Compared to running A-C alone, running all groups together gave me more genes that are differentially expressed, though the top 50 seems to overlap a bit. Can anyone enlighten me on why that is?

Thanks!