Summer project
Bioinformatics Specialization · Cohort 2024
Quantifying the Effect of Joint Calling
Joint variant calling improves the accuracy of variant detection by analyzing genetic data across multiple samples simultaneously. It leverages population statistics and biological assumptions about inheritance to enhance variant detection accuracy. A previous experiment measured the benefits of joint calling and demonstrated that increasing the size of the cohort improves variant sensitivity, with benefits plateauing around 600 samples (Gauthier et al., 2016, ASHG). The goal of this research was to reproduce this experiment using modern genetic data, measure the benefits of joint calling, and identify the minimum sample size required to achieve optimal performance. Additionally, we explored how varying population structures within the cohort affect the accuracy of joint variant calling. We analyzed genotypic changes by creating cohorts of increasing size from the 1000 Genomes dataset, ensuring the inclusion of the well-established HG002 sample in each cohort. We performed joint calling using the GenotypeGVCFs tool. The HG002 sample was then extracted from the joint VCF files and benchmarked using VCFeval.
With KAUST Academy
- Cohort 2024
Broad Institute of MIT and Harvard · Bioinformatics Summer Internship
Want your entry updated or removed? Contact us.