Evaluation of variant calling tools for detection of SNVs in BRCA1 and BRCA2 genes in patients from the Institute of Oncology and Radiology of Serbia

Abstract

Serbia has one of the world’s highest incidences and mortality rates of ovarian cancer. Germline or somatic mutations in BRCA1 and BRCA2 genes, such as single nucleotide variants (SNVs), indels, insertions, deletions, commonly lead to development of breast and ovary cancer. Targeted therapy with PARP inhibitors is the current standard of care for serous epithelial BRCA-mutated ovarian cancer and depends on the accurate detection of mutations in these genes. In this study, a subset of patient specimens from Institute of Oncology and Radiology were sequenced on MiSeq Illumina sequencer, raw data were analysed bioinformatically, which included checking quality control of raw FASTQ sequences, trimming, mapping them on reference genome(hg19), target coverage quality control and variant calling. We tested various variant calling tools including Mutect2, GATK HaplotypeCaller, FreeBayes, VarDict and MuSe callers. We evaluated the relative performance- concordance rate, false positive and false negative rates between the callers for SNV/indel detection in BRCA1 and BRCA2 genes.