ATACUT -- Pipeline for caling Tn5 Cutsite peaks

Common practice is to call ATAC-seq peaks from fragment pileups, but fragments represent 2 independent Tn5 cut events - one at either end of the fragment - and can obscure biologically relevant details about chromatin accessibility.

This is a simple Nextflow pipeline to...

1. Convert Bam files to bed
2. Call peaks on bed files with Macs2 and settings to shift peak calling around cut sites instead of fragments
3. Generate bigWig coverage files with small estimated cut site fragments
4. Apply z-normalization to bigwigs

Warning

I initially wrote this workflow as part of a larger analysis and deployed on our university HPC. I plan to get this set up for running on GCP as well - but for now it should be ready to go for systems using slurm or LSF.