Allelic_variant_call_pipeline

Pipeline for allele-specific variant calling from RNAseq. This pipeline has been optimized to work with ERCC exRNA samples on BRL-BCM HPC.

Prerequisite tools and databases

1. Construction of the diploid genomes requires the following dependencies:

• gnomAD frequency database hg38_gnomad30_genome.txt available at Annovar Main Package
• vcf2diploid
• Crossmap

2. The variant calling pipeline requires the following dependencies:

• STAR, gatk, picard, samtools, BEDTools
• Hg38 reference file,reference annotation file and known variants sites files such as 1000G_phase1.snps.high_confidence.hg38.vcf.gz, Mills_and_1000G_gold_standard.indels.hg38.vcf.gz, all available at GATK resource bundle.

Running the pipeline

Step 1: Keep only genome regions covered by ERCC samples

1. getBedCoveragePerStudy.pbs.txt: generate unioned coverage depth bedfile across ERCC studies
2. getSumBedCoveragePerStudy.pbs.txt: utilize unionbedgByStudy.R to sumarize covered regions by study
3. maskRefGenome.pbs.txt: mask uncovered genome regions.
4. Convert unioned ERCC genome coverage bedgraphs from version hg19 to hg38

Step 2: Construct (artificial) diploid genome

Step 3: Call extracellularly expressed variants

ppl_vCall_ref_Hr4Rm47Prc4.pbs: calls extracellularly expressed variants with minimum reference bias.