Numt identification and removal with RtN!

• DOI: 10.1093/bioinformatics/btaa642
• Publication Date: 2020-07-18T11:10:07Z
• Authors: August E Woerner; Jennifer Churchil...; Utpal Smart; Bruce Budowle
• Keywords: Cell Nucleus; 0301 basic medicine; 03 medical and health sciences; Genome, Mitochondrial; High-Throughput Nucleotide Sequencing; Sequence Analysis, DNA; Software

Abstract

Motivation Assays in mitochondrial genomics rely on accurate read mapping and variant calling. However, there are known and unknown nuclear paralogs that have fundamentally different genetic properties than that of the mitochondrial genome. Such paralogs complicate the interpretation of mitochondrial genome data and confound variant calling.

Results Remove the Numts! (RtN!) was developed to categorize reads from massively parallel sequencing data not based on the expected properties and sequence identities of paralogous nuclear encoded mitochondrial sequences, but instead using sequence similarity to a large database of publicly available mitochondrial genomes. RtN! removes low-level sequencing noise and mitochondrial paralogs while not impacting variant calling, while competing methods were shown to remove true variants from mitochondrial mixtures.

Availability and implementation https://github.com/Ahhgust/RtN

Supplementary information Supplementary data are available at Bioinformatics online.