kosarhooshmand / rnaseq-based-snp-calling

✒️ RNAseq-SNP-calling

My Guthub repo include pipelines for large-scale SNP detection from RNA-seq data of different tissues per phenotype. Our pipeline mapped raw sequence reads against the genome using STAR aligner (2-pass method). Uniquely mapped reads were then pre-processed using GATK Best-Practices pipeline for RNA-seq data. This was followed by variant detection processes, and vigorous filtering of false-positive calls.

⚙️Technologies & Tools

🔗 Quick Start

1. Mapping RNA-seq data to the reference using STAR

In this resaerch, the STAR pipeline was written in Workflow Description Language (WDL).

• Dependencies

a. Docker

b. Cromwell

c. Java

❗ For running paired-star-align.wdl you also need to build an index from Gencode-Annotation-File using STAR and save them in a seprated folder.

• Run the workflow directly by executing the following commands on your terminal:

java -Dconfig.file=application.conf -jar cromwell-55.jar run paired-star-align.wdl -i paired-star-align.json

2. Pre-processing measures using picard-tools

Use the Pre-processing-picard.sh pipeline for this step.

• Required tools

a. picard.jar

• Required data

a. Bam files produced from STAR aligner

b. Homo_sapiens_assembly38.dict

3. Variant calling steps using GATK Best-Practice pipeline

Use the Variant-calling-GATK.sh pipeline for this step.

• Required tools

a. GATK Best Practices

• Required data

a. Output of the last step of pre-processing pipeline

b. Homo_sapiens_assembly38.fasta

c. Homo_sapiens_assembly38.dbsnp138

d. Homo_sapiens_assembly38.dbsnp138.indexed

e. Homo_sapiens_assembly38.known-indels

f. Homo_sapiens_assembly38.known_indels.indexed

g. wgs_calling_regions.hg38.interval_list

4. Merge VCFs & Filteration

Use the VCFs-merge-filter.sh pipeline for this step.

• Required tools

a. bcftools

b. tabix

c. GATK Best Practices

• Required data

a. A list of selected VCFs