Clairvoyante

Introduction

Clairvoyante is a deep neural network based variant caller.

For more information, please check:

Docker hub: https://hub.docker.com/r/lifebitai/clairvoyante

Home page: https://github.com/aquaskyline/Clairvoyante

Versions

• 1.02

Commands

• clairvoyante.py

Module

You can load the modules by:

module load biocontainers
module load clairvoyante

Example job

Warning

Using #!/bin/sh -l as shebang in the slurm job script will cause the failure of some biocontainer modules. Please use #!/bin/bash instead.

To run clairvoyante on our clusters:

#!/bin/bash
#SBATCH -A myallocation     # Allocation name
#SBATCH -t 1:00:00
#SBATCH -N 1
#SBATCH -n 1
#SBATCH --job-name=clairvoyante
#SBATCH --mail-type=FAIL,BEGIN,END
#SBATCH --error=%x-%J-%u.err
#SBATCH --output=%x-%J-%u.out

module --force purge
ml biocontainers clairvoyante

cd training
clairvoyante.py callVarBam \
   --chkpnt_fn ../trainedModels/fullv3-illumina-novoalign-hg001+hg002-hg38/learningRate1e-3.epoch500 \
   --bam_fn ../testingData/chr21/chr21.bam \
   --ref_fn ../testingData/chr21/chr21.fa \
   --bed_fn ../testingData/chr21/chr21.bed \
   --call_fn chr21_calls.vcf \
   --ctgName chr21