Short read mapping pipeline pig

The latest short-read mapping pipeline for the pig project is based on a Python3 script that creates a shell script that can subsequently be executed from the command line or submitted to the cluster using SLURM. The latest version of the Python3 script can be found at GitHub.

Prerequisites

Data sources

• path to sequence archives /lustre/nobackup/WUR/ABGC/shared/Pig/ABGSA/ (for pig only)
• access to ABGSA meta-database, currently hosted at scomp1095.wurnet.nl, database='ABGSAschema')
• path to reference genome, including index for BWA /lustre/nobackup/WUR/ABGC/shared/Pig/Sscrofa_build10_2/Ensembl72/Sus_scrofa.Sscrofa10.2.72.dna.toplevel.fa

Hardcoded paths

All paths to data and software are currently hardcoded. This is done for transparency (hardcoded==explicit). Hardcoded paths do require work however, when migrating to new environment. Currently contemplated to switch to using environment variables.

• bwa 5.9 /cm/shared/apps/WUR/ABGC/bwa/bwa-0.5.9/
• bwa 7.5 /cm/shared/apps/WUR/ABGC/bwa/bwa-0.7.5a/
• samtools 1.19 /cm/shared/apps/WUR/ABGC/samtools/samtools-0.1.19/
• samtools 1.12 /cm/shared/apps/WUR/ABGC/samtools/samtools-0.1.12a/
• picard /cm/shared/apps/WUR/ABGC/picard/picard-tools-1.93/
• GATK /cm/shared/apps/WUR/ABGC/GATK/GATK2.6/
• Mosaik /path/to/mosaik/ref.dat
• Mosaik Jump Library /path/to/mosaikjump/ref.j15
• dbSNPfile=reffolder+'/dbSNP/dbSNP.vcf'
• gatk_gvcf /cm/shared/apps/WUR/ABGC/GATK/GATK_gVCFmod/
• gvcftools /cm/shared/apps/WUR/ABGC/gvcftools/gvcftools-0.16/bin/
• helper scripts /cm/shared/apps/WUR/ABGC/abgsascripts/
• path to sqlite db path (for cow) /path/to/sqlite/Bulls1000/
• Variant Effect Predictor (VEP) /cm/shared/apps/WUR/ABGC/variant_effect_predictor/VEP231213/

Present in the working directory

• cow_schema.db (SQLite db for 1000 Bulls project - for cow only at the moment).

Basic execution

<source lang='bash'> (virtenv)[megen002@nfs01 rundir]$ python ABGC_mapping_v2.py -i LW22F04 -a /lustre/nobackup/WUR/ABGC/shared/Pig/ABGSA/ -r /lustre/nobackup/WUR/ABGC/shared/Pig/Sscrofa_build10_2/Ensembl72/Sus_scrofa.Sscrofa10.2.72.dna.toplevel.fa -t 4 </source> The code should produce the following shell script, ready for execution with SLURM.

Automated runfile creation

<source lang='bash'> mysql -u ABGSAuser -h scomp1095.wurnet.nl -p ABGSAschema -e 'select ABG_individual_id from ABGSAschema_main where archive_name like "ABGSA0%" group by ABG_individual_id' >list.txt FILES=`cat list.txt` for ID in $FILES; do python ABGC_mapping_v2.py -i $ID -a /lustre/nobackup/WUR/ABGC/shared/Pig/ABGSA/ -r /lustre/nobackup/WUR/ABGC/shared/Pig/Sscrofa_build10_2/Ensembl72/Sus_scrofa.Sscrofa10.2.72.dna.toplevel.fa -t 4; done </source>

See also

• Animal Breeding & Genomics Sequence Archives
• 1000 Bulls @ABGC implementation of the pipeline

External links

NGStools page on GitHub