NVIDIA-Clara-Parabricks-Somatic-Variant-Calling-on-AWS

AWS Blog (Somatic Variant Calling )

Data download required to run NVIDIA Clara Parabricks Somatic Variant Calling Pipeline.

Download publicly available BAM files using wget

WGS Normal sample (50X)

BAM: wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WGS/WGS_EA_N_1.bwa.dedup.bam

File size: 95GB

BAM INDEX: wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WGS/WGS_EA_N_1.bwa.dedup.bai

WGS Tumor sample (50X)

BAM: wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WGS/WGS_EA_T_1.bwa.dedup.bam

File size: 107GB

BAM INDEX: wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WGS/WGS_EA_T_1.bwa.dedup.bai

Download 100x Whole exome dataset from SEQC2 hosted at NCBI:

WES Normal Sample (80X)

• wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WES/WES_EA_N_1.bwa.dedup.bam
• wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WES/WES_EA_N_1.bwa.dedup.bai

WES Tumor Sample (100X)

• wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WES/WES_EA_T_1.bwa.dedup.bam
• wget https://ftp-trace.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/WES/WES_EA_T_1.bwa.dedup.bai

Downloading GRCh38 reference genome along with known sites from GDC/SEQC2, these files will be used for alignment and variant calling.

The reference files are hosted at GDC/NCI website. (https://gdc.cancer.gov/about-data/gdc-data-processing/gdc-reference-files)

Fasta reference files: GRCh38.d1.vd1.fa.tar.gz

wget https://api.gdc.cancer.gov/data/254f697d-310d-4d7d-a27b-27fbf767a834 -O GRCh38.d1.vd1.fa.tar.gz
tar -xvzf GRCh38.d1.vd1.fa.tar.gz

BWA index reference files: GRCh38.d1.vd1_BWA.tar.gz

wget https://api.gdc.cancer.gov/data/25217ec9-af07-4a17-8db9-101271ee7225 -O GRCh38.d1.vd1_BWA.tar.gz
tar -xvzf GRCh38.d1.vd1_BWA.tar.gz 

dbSNP file to use : dbSNP146

wget ftp://[email protected]:21/bundle/hg38/dbsnp_146.hg38.vcf.gz

InDel file from 1k genome to use for known sites : 1000 genome Phase 3

wget https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/other_mapping_resources/ALL.wgs.1000G_phase3.GRCh38.ncbi_remapper.20150424.shapeit2_indels.vcf.gz
## Please edit line 22 of this file add missing double quote before dbSNP
   Description=“dbSNP ssID of the allele” 
   ##INFO=<ID=ssID,Number=A,Type=String,Description=dbSNP ssID of the allele">
## Please edit line 42 change “POS=POS-1” to “POS_POS-1”
   ##INFO=<ID=POS=POS1,Number=0,Type=Flag,Description="POS has been adjusted due to missing REF in NCBI VCF file">

Curated InDel file from: Mills_and_1000G_gold_standard.indels

wget https://ftp.1000genomes.ebi.ac.uk/vol1/ftp/technical/reference/GRCh38_reference_genome/other_mapping_resources/Mills_and_1000G_gold_standard.indels.b38.primary_assembly.vcf.gz

Before starting the analysis run, please install following software’s will be handy for some data processing:Download accessory tools

  cd /mnt/disks/local
  ## Downlaod and install GATK 
  wget https://github.com/broadinstitute/gatk/releases/download/4.2.0.0/gatk-4.2.0.0.zip
  unzip gatk-4.2.0.0.zip
  ## Install samtools
  sudo apt install samtools
  ## Install Tabix
  sudo apt install tabix
  ## index fasta file
  samtools faidx Refs/GRCh38.d1.vd1.fa
  ## create dictionary file from fasta: 
  gatk-4.2.0.0/gatk CreateSequenceDictionary R=/home/ubuntu/Refs/GRCh38.d1.vd1.fa O=/home/ubuntu/Refs/GRCh38.d1.vd1.dict

Command line settings to run the Somatic Variant callers

#### Caller1: Mutect2 variant calling ($GPU=2,4,8), here we used $GPU=4
pbrun mutectcaller --ref Refs/GRCh38.d1.vd1.fa \
--in-tumor-bam WGS_EA_T_1.bwa.dedup.bam \
--in-normal-bam WGS_EA_N_1.bwa.dedup.bam \
--in-tumor-recal-file WGS_EA_T_1.bwa.dedup.recal.txt \
--in-normal-recal-file WGS_EA_N_1.bwa.dedup.recal.txt \
--out-vcf WGS_EA_TN-1-mutect2.vcf \
--tumor-name WGS_EA_T_1 \
--normal-name WGS_EA_N_1
--num-gpus $GPU \


#### Caller2: MuSE
##### Step1: MuSE call
pbrun muse --ref Refs/GRCh38.d1.vd1.fa \
--in-tumor-bam WGS_EA_T_1.bwa.dedup.bam \
--in-normal-bam WGS_EA_N_1.bwa.dedup.bam \
--out-file WGS_EA_TN-1-muse.call \
--num-threads 16 --mode call
##### Step2: MuSe sump 
pbrun muse --ref Refs/GRCh38.d1.vd1.fa \
--in-callfile WGS_EA_TN-1-muse.call.MuSE.txt \
--datatype G \
--out-vcf WGS_EA_TN-1-muse.call.MuSE.vcf \
--num-threads 16 --mode sump \
--in-dbsnp Refs/dbsnp_146.hg38.vcf.gz


#### Caller3: Somatic Sniper
pbrun somaticsniper --ref Refs/GRCh38.d1.vd1.fa \
--in-tumor-bam WGS_EA_T_1.bwa.dedup.bam \
--in-normal-bam WGS_EA_N_1.bwa.dedup.bam \
--min-mapq 1 --no-gain --no-loh --out-format vcf \
--out-file WGS_EA_TN-1-somaticsniper.vcf \
--num-threads 16 \


#### Caller4: LoFreq
pbrun lofreq --ref Refs/GRCh38.d1.vd1.fa \
--in-tumor-bam WGS_EA_T_1.bwa.dedup.bam \
--in-normal-bam  WGS_EA_N_1.bwa.dedup.bam \
--output-dir pbrun_LoFreq \
--num-threads 4 --num-gpus $GPU \
--in-dbsnp-file Refs/dbsnp_146.hg38.vcf.gz