Comparison of Two Variant Analysis Programs for Next-Generation Sequencing Data of Whole Mitochondrial Genome

Abstract

Background With advance of next-generation sequencing (NGS) techniques, the need for mitochondrial DNA analysis is increasing not only in the forensic area, but also in medical fields. Methods Two commercial programs, Converge Software (CS) and Torrent Variant Caller for variant calling of NGS data, were compared with a considerable amount of sequence data of 50 samples with a homogeneous ethnicity. Results About 2,300 variants were identified and the two programs showed about 90% of consistency. CS, a dedicated analysis program for mitochondrial DNA, showed some advantages for forensic use. By additional visual inspection, several causes of discrepancy in variant calling results were identified. Application of different notation rules for mitochondrial sequence and the minor allele frequency close to detection threshold were the two most significant reasons. Conclusion With prospective improvement of each program, researchers and practitioners should be aware of characteristics of the analysis program they use and prepare their own strategies to determine variants.