Hi all! I'm new to RNA sequencing and have some questions about an experimental design and how this might affect my downstream analysis.

I want to purify cells from mice, treat them with different ligands, and then perform RNA-seq to curate a list of genes with fold changes greater than 2.

My biological question is what genes display large fold-changes under these different ligand conditions?

I already have a list of gene categories that I'm interested in, so I have some idea of what I'm looking for.

My current experimental plan is the following:

• purify cells from 3 mice and pool them together to get rid of variations between mice
• set up experiment where each treatment group gets 2 replicates
• run experiment and prepare total RNA for all samples at the same time

My questions are the following:

1. I need to pool cells together in order to have enough for the entire sequencing experiment. If I set up two replicates for each treatment group using the pool of cells from 3 mice, can I call those biological replicates or are they some variation of technical replicates? It's a bit awkward to think about because it's like trying to convert in-vivo to in-vitro.

2. If I get rid of one treatment group, I might be able to divide the cells from one mouse across all of the treatment groups. In this case, each one of my replicates would correspond to a separate mouse. These, I'm assuming, would be true biological replicates. Is this advisable?

3. I've talked with colleagues about this experiment and they have recommended against any under 3 biological replicates. If the goal of this experiment is simply to curate a list of genes with high fold-changes relative to baseline and not necessarily to publish the data, would 2 replicates be ok?

4. Is there such thing as an analysis that focuses on fold-changes rather than p-values, or are the two always linked?

Thank you!!