ATAC peak calling

[SichongP]

in functional-annotation/Rules/CallPeaks.smk, line 180,

No -B option used in macs2 callpeak command, which resulted in no treat_pileup.bdg files.

Then in rule fold_enrichment (line 199):

pipeline attempts to call macs2 bdgcmp which uses bedGraph files generated in peak calling. And since there are no such files, this results in an error that shuts down pipeline.

And since we don't run "control" for ATAC, the bdgcmp call seems incorrect. Am I missing something?

[kernco]

All I've done right now for the ATAC-seq part of the pipeline is just copied the commands from Michelle's scripts into the pipeline, but I haven't really tested it yet. There's a lot of downstream analyses past the basic alignment and peak calling that it's going to try to do but that's all designed for ChIP-seq data. Some of it might be worthwhile but other stuff it probably shouldn't do or it should do it differently. Right now I would just use the pipeline for getting the peak calls, then do your own analyses from those.

[SichongP]

I see. Thank you!