Hi All,

Longtime lurker hoping to get guidance on a package that may best fit my RNA-chromatin data. The technique is similar to: https://www.nature.com/articles/nbt.3968

My library generates pairs of RNA in close proximity to DNA. I have replicates for two time points and generated an interaction score for any RNA to 1KB bins of the genome. The scores are fold enrichment over background and normalized to library size already. Below is an example table. As one approach I collapsed the bins to protein coding regions and averaged the interaction score.

My question: Is there a package that I can use (maybe DiffBind) to generate p-values to demonstrate that RNA-1 significantly interacts with Gene 1 3 fold higher at Day 0 vs Day 2. Additionally RNA-1 interacts 4 fold higher with Gene 3 at Day 2 vs Day 0.

Any advice, packages I could use would be greatly appreciated. Thank you!

If you wanted to use DESeq2 for this, the key would be to input the original counts as the counts matrix. Library size is corrected for internally, and you don't divide this out before passing the data to DESeq2.