Benchmarking whole exome sequencing pipeline for predicting pathogenic variants of significance

Abstract

Benchmarking whole exome pipelines is crucial for evaluating and comparing their performance in variant calling and clinical significance annotation. It enables researchers and clinicians to assess the accuracy, sensitivity, and specificity of different pipelines and identify the most effective and reliable ones. In this study, we evaluated and compared the performance of our in-house consensus exome pipeline with a widely recognized gold standard Genome Analysis Toolkit (GATK) pipeline. Four datasets were used for evaluation, three 1000 Genome Project (1KGP) datasets and one Prostate cancer (PCa) Sample. The consensus pipeline consistently demonstrated a higher average transition-to-transversion (Ti/Tv) ratio, indicating enhanced precision in identifying single nucleotide variant (SNV) calls. This suggests that the consensus pipeline excels in effectively discerning true genetic variations from sequencing artefacts, particularly in the context of exome sequencing. Additionally, the pipeline exhibited increased sensitivity in detecting pathogenic and likely pathogenic variants in the PCa sample, further highlighting its efficacy in identifying clinically relevant variants. We also conducted a trio exome analysis with the use of two trio pipelines,viz. VarScan Trio and GATK joint calling pipelines on our erstwhile Congenital Pouch Colon (CPC) samples from our rare disease cohort which we published earlier and found that the GATK predicted a significantly higher number of variants compared to VarScan. Our study demonstrates a large potential for trio-variant calling analysis paving the way for precision medicine.