High-Performance Computing at the NIH
GitHub YouTube @nih_hpc RSS Feed

TVC is the standalone Torrent Variant Caller, part of the Ion Torrent Suite.

There are multiple versions of TVC available. An easy way of selecting the version is to use modules. To see the modules available, type

module avail TVC

To select a module, type

module load TVC/[ver]

where [ver] is the version of choice.

Environment variables set:

On Helix

Sample session:

$ module load TVC
$ variant_caller_pipeline.py -i /path/to/input.bam -r /path/to/reference/file.fasta -N [NUMTHREADS]

NOTE: The default number of threads ([NUMTHREADS]) is 12.

Batch job on Biowulf

Create a batch input file (e.g. TVC.sh), which uses the input file 'input.bam'. For example:

module load TVC
variant_caller_pipeline.py -i /path/to/input.bam -r /path/to/reference/file.fasta -N [NUMTHREADS]

Submit this job using the Slurm sbatch command.

$ sbatch --cpus-per-task=12 --mem=20g TVC.sh
Swarm of Jobs on Biowulf

Create a swarmfile (e.g. TVC.swarm). For example:

mkdir outdir1 ; variant_caller_pipeline.py -i /path/to/input1.bam -r /path/to/reference/file.fasta -N 8 -o outdir1
mkdir outdir2 ; variant_caller_pipeline.py -i /path/to/input2.bam -r /path/to/reference/file.fasta -N 8 -o outdir2
mkdir outdir3 ; variant_caller_pipeline.py -i /path/to/input3.bam -r /path/to/reference/file.fasta -N 8 -o outdir3
mkdir outdir4 ; variant_caller_pipeline.py -i /path/to/input4.bam -r /path/to/reference/file.fasta -N 8 -o outdir4

NOTE: By default, all output is written in the current working directory. When run in a swarm, it is ideal to create a directory for the output and redirect that output using the -o option. This will prevent overwriting the same file by indendent swarm subjobs.

Submit this job using the swarm command.

$ swarm -f TVC.swarm -t 8 -g 20
Interactive job on Biowulf

Once an interactive job has been started,

sinteractive --cpus-per-task=8 --mem=20g

load the module and run the command.