DRAGoN: a robust pipeline for analyzing DRUG-seq datasets

Motivation Existing bioinformatics pipelines to process DRUG-seq datasets have limited flexibility and can have difficulty analyzing current datasets without requiring excessive computational time or memory. Results Here we describe an alternative, DRAGoN, which is fast, robust, and performs as well as or better than competing pipelines on key benchmarks without sacrificing accuracy. This is accomplished primarily via a preliminary demultiplexing step that facilitates the parallelization of the pipeline as well as the collection of per-well statistics that assist with quality control. DRAGoN provides the user maximum flexibility with respect to filtering, alignment, counting, and downsampling, and it efficiently collapses UMIs. Availability and implementation DRAGoN is a Nextflow pipeline that utilizes open-source software alongside custom C ++ programs and Python scripts. It is freely available at https://github.com/MSDLLCPapers/DRAGoN