razmia02/targeted-exome-seq-analysis: End-to-end workflow for analyzing exome seq data. · GitHub

This project analyzes paired-end targeted exome sequencing data generated by Illumina platform. The sequencing was performed using custom gene panels. The workflow focuses on identifying and characterizing germline variants (SNPs and Indels) using the GATK Best Practices pipeline, followed by filtering and annotation to generate high-confidence variant calls within the targeted regions. Note Raw sequencing data is not included in this repository as it belongs to the clinical laboratory and contains patient-derived samples. Output files (filtered and annotated VCFs) are included in final_results/ to demonstrate pipeline outputs.

Identify clinically relevant germline variants in cancer predisposition genes from targeted exome seq data, following GATK Best Practices Pipeline.

1. Quality Control

   • Tool: FastQC
   • Purpose: Assess base quality, GC content, adapter contamination of raw reads.

2. Read Alignment

   • Tool: BWA-MEM
   • Align reads to the reference genome (Hg19).

3. Post-Alignment Processing

   • Tools: samtools, GATK MarkDuplicatesSpark
   • Steps: Flag duplicated reads, remove them and convert SAM to BAM.

4. Base Quality Score Recalibration (BQSR)

   • Tool: GATK BaseRecalibrator, GATK ApplyBQSR
   • Recalibrate the base quality scores.

5. Variant Calling

   • Tool: GATK HaplotypeCaller
   • Restricted to targeted regions using the BED file.

6. Variant Filtration

   • Tool: GATK SelectVariants, GATK VariantFiltration
   • Select SNPs and Indels and further filter the VCF by applying hard filters.

7. Variant Annotation

   • Tools: GATK VariantAnnotator, bcftools annotate
   • Annotate variants with dbSNP IDs and gene names from BED file. Note: Only selected result files and read samples are included in this repo due to the large size of other data files.

Some variants observed within the samples are as follows.

No pathogenic or likely pathogenic variants were identified in this set of samples. All detected variants are classified as benign or variants of uncertain significance (VUS) in ClinVar, consistent with a germline screening result in unaffected or low-risk individuals.

• Single-sample pipeline: The current bash script processes one sample at a time. A production-grade extension would implement joint genotyping across a cohort using GATK's genomic GVCF mode.
• Annotation depth: Annotation is currently limited to dbSNP IDs and gene names. Integration with a clinical annotation tool (e.g., VEP with ClinVar, OMIM, and SpliceAI plugins) would provide richer clinical context.
• Pipeline management: Migrating from a monolithic bash script to a workflow manager (Snakemake or Nextflow) would improve reproducibility, parallelisation, and portability across HPC environments.
• Variant interpretation: Automated ACMG classification (e.g., via InterVar) is a planned addition for clinical reporting.

To run this targeted exome sequencing pipeline locally, open your terminal, clone the repository, and navigate into the project directory:

git clone https://github.com/razmia02/targeted-exome-seq-analysis.git)
cd targeted-exome-seq-analysis

# Create the environment from the configuration file
conda env create -f env.yaml

# Activate the environment
conda activate exome_seq

Before running the pipeline, make sure to download and place your reference genome (human_g1k_v37.fasta) and known variant sites VCF files (dbsnp_138.b37.vcf, Mills_and_1000G_gold_standard.indels.b37.vcf) into the ref_files/b37/ directory.

chmod +x script.sh
./script.sh

targeted-exome-seq-analysis/
├── reads/                      # Raw & processed reads
├── script.sh                   # shell script for analysis
├── final_results/               # Raw, processed & final annotated indels & snps
└── README.md                  # Project documentation
└── env.yaml                    # Conda environment

• About GATK Best Practices
• Data Pre-processing
• Read Group
• About BQSR
• HaplotypeCaller tool
• VariantAnnotator tool