Nextflow workflow for analyzing CRISPR screen NGS data
The purpose of this workflow is to analyze NGS data generated from a CRISPR screen, which consists of two groups of FASTQ files. One of the groups of FASTQ files is considered a 'control', meaning that it was generated from a set of CRISPR sgRNA guides which did not undergo an experimental selection process. The other group of FASTQ files are the 'treatment', indicating that they did undergo some experimental selection. The purpose of this analysis is to quantify and summarize the degree of that experimental selection process on each of the sgRNA guides in the library. In other words, to determine what the biological impact of knocking down each particular gene was in this biological context.
In addition to the two groups of FASTQ files, the user must provide a 'library'
file which describes the set of sgRNA guides used in the screen, along with the
genes which each guide targets. Documentation describing this library file can
be found here. The file format
expected by this particular workflow is CSV, with the three column headers named:
guide,sgrna,gene.
After generating NGS data from the sgRNA libraries used in this screen, the user must specify the best way to trim adapter sequences away from the sgRNA guides. The two options available using this workflow are (a) trimming a fixed number of bases from the 5' of each read, and then trimming down to a fixed number of bases, or (b) using a known adapter sequence to remove 5' adapters, and then trimming down to a fixed number of bases. In either case, the reads which are output by the trimming step will all have the same length (which should correspond to the length of the sgRNA guides).
To trim a fixed number of bases from the 5' of each read, use the trim_5_prime
parameter. To trim a known 5' adapter sequence, use the adapter_5_prime parameter.
Details on sequence-based 5' adapter trimming can be found
here.
Before using this pipeline, the user must install Nextflow and configure Nextflow to use the appropriate set of computational resources which are available. The software dependencies for this pipeline are managed using Docker images (instead of Conda environments), and so support must be added for Docker or Singularity as appropriate. The usage documentation listed below assumes a funcational Nextflow configuration present on the system used to run the workflow.
Usage:
nextflow run FredHutch/crispr-screen-nf
Required Arguments:
--treatment_fastq Path to FASTQ data for treatment samples
--control_fastq Path to FASTQ data for control samples
Multiple files can be specified with wildcards and commas, e.g.
/path/to/inputs/A/*.fastq.gz,/path/to/inputs/B/*.fq.gz
--library Text file describing sgRNA library
As described at https://sourceforge.net/p/mageck/wiki/input/
--output Path to output directory
--insert_length Length of sgRNA guides (default: 20)
--trim_5_prime Number of bases to trim from the 5' of each read (default: 0)
--adapter_5_prime (optional) Sequence of 5' adapter to be trimmed from each read
--organism Organism string provided for MAGeCK-Flute (default: hsa)
--scale_cutoff Parameter 'scale_cutoff' for MAGeCK-Flute (default: 1)
--gmt Pathway GMT File
--depmap_effect "https://ndownloader.figshare.com/files/20234073" - Effect File Can't Download From Different Region
--depmap_samples "https://ndownloader.figshare.com/files/20274744" - Sample File Can't Download From Different Region