Page Comparison

For this exercise we will use the epi2me-labs/wf-human-variation pipeline. Find information on the pipeline at https://labs.epi2me.io/workflows/wf-human-variation/

...

Approximate run time: Variable depending on whether it is targeted sequencing or whole genome sequencing, as well as coverage and the individual analyses requested. For instance, a 90X human sample run (options: --snp --sv --mod --str --cnv --phased --sex male) takes less than 8h with recommended resources.

NOTE: in contrast to the nf-core/sarek pipeline that we used in session 2, the epi2me-labs/wf-human-variation pipeline runs in ‘local’ mode (needs large amount of CPUs and RAM memory), while the nf-core/sarek pipeline will use a ‘pbspro’ mode, where the pipeline will submit individual jobs to the HPC cluster and define the CPUs and memory for each task individually.

...

#!/bin/bash -l
#PBS -N run3_epi2me_WF-HV
#PBS -l select=1:ncpus=16:mem=32gb
#PBS -l walltime=48:00:00
#PBS -m abe

module load java

export NXF_OPTS='-Xms1g -Xmx4g'
cd $PBS_O_WORKDIR

nextflow run epi2me-labs/wf-human-variation \
    -r v2.2.0 \
    -profile singularity \
    --bam "/training/ONTvariants/runs/run2_mapping/SRR17138639_mapped_chr20.bam" \
    --basecaller_cfg "dna_r10.4.1_e8.2_400bps_hac_prom" \
    --mod \
    --ref "/training/ONTvariants/data/chr20.fasta" \
    --sample_name "SRR17138639" \
    --snp \
    --sv \
    --bam_min_coverage 15 \
    --out_dir "results"

Note:

• Line 1: Defines that the script is a bash script.

• Lines 2-5: Are commented out with “#” at the beginning and are ignored by bash, however, these PBS lines tell the scholar (PBS Pro) the name of the job (line 2), the number of CPUs and RAM memory to use (line 3), the time to run the script (line 4) and report if there are any errors (line 5).

• Line 7: load java required to run cextflow pipelines

• Line 8: assign up to 4GB memory for the nextflow initial script to use

• Line 9: Tells the job to run on the current directory.

• Lines 11-22: Parameters to run the epi2me-labs/wf-human-variation pipeline (refer above for details on each parameter)

Submit the QC job to the HPC cluster:

qsub launch_ONTvariants_mapping.pbs

Monitor the progress of the job:

qjobs