A bioinformatics pipeline for estimating mitochondrial DNA copy number and heteroplasmy levels from whole genome sequencing data

Abstract

Mitochondrial diseases are a heterogeneous group of disorders that can be caused by mutations in the nuclear or mitochondrial genome. Mitochondrial DNA (mtDNA) variants may exist in a state of heteroplasmy, where a percentage of DNA molecules harbor a variant, or homoplasmy, where all DNA molecules have the same variant. The relative quantity of mtDNA in a cell, or copy number (mtDNA-CN), is associated with mitochondrial function, human disease, and mortality. To facilitate accurate identification of heteroplasmy and quantify mtDNA-CN, we built a bioinformatics pipeline that takes whole genome sequencing data and outputs mitochondrial variants, and mtDNA-CN. We incorporate variant annotations to facilitate determination of variant significance. Our pipeline yields uniform coverage by remapping to a circularized chrM and by recovering reads falsely mapped to nuclear-encoded mitochondrial sequences. Notably, we construct a consensus chrM sequence for each sample and recall heteroplasmy against the sample's unique mitochondrial genome. We observe an approximately 3-fold increased association with age for heteroplasmic variants in non-homopolymer regions and, are better able to capture genetic variation in the D-loop of chrM compared to existing software. Our bioinformatics pipeline more accurately captures features of mitochondrial genetics than existing pipelines that are important in understanding how mitochondrial dysfunction contributes to disease.

Mitochondrial DNA control region typing from highly degraded skeletal remains by single‐multiplex next‐generation sequencing

“…Moreover, homopolymeric Cstretch regions located at nucleotide positions 16 184-16 193 (amplicon 2) in HVI and at positions 303-315 in HVII (amplicon 8) are highly prone to present heteroplasmy produced by the insertion/deletion of cytosines [45]. The presence of homopolymeric C-stretch generates low quality reads, reducing the number of reads that pass quality filters, and the alignment of these regions is usually poor [46,47]. Accordingly, the low performance of amplicons 2 and 8 can also be related to homopolymeric C-stretch.…”

Deleterious heteroplasmic mitochondrial mutations increase risk of overall and cancer-specific mortality

“…Mitochondrial heteroplasmic mutations are composed of a combination of inherited and somatic mutations 7 . We would therefore expect that heteroplasmic mutations would increase with age, as has been previously shown 9,17 , as well as with exposure to mutagenic chemicals such as tobacco. Indeed, we observed a significant increase in heteroplasmic SNVs both as a function of age and smoking status (Fig.…”

“…We further excluded heteroplasmic variants at poly-C homopolymer regions on the mitochondrial chromosome and excluded INDELs. INDELs are often found at homopolymer regions and due to the nature of these regions, are challenging to accurately call heteroplasmies 9 . Nuclear-encoded mitochondrial sequences (NUMTs) can contribute to false positive mtDNA heteroplasmy calls.…”

“…We developed a bioinformatics pipeline, MitoHPC, to accurately measure mtDNA SNVs in large WGS datasets 9 . MitoHPC’s key feature is that it constructs a consensus mitochondrial sequence for each individual, which allows for more accurate read mapping and for measuring heteroplasmy against an individual’s unique mitochondrial genome.…”

Quantifying constraint in the human mitochondrial genome

“…Mitochondrial heteroplasmy was identified from whole genome sequencing (WGS) data from the UK Biobank, a large population study of people from the United Kingdom aged 40-69 years 72 , as described previously 49 . In brief, MitoHPC 73 was used to call heteroplasmic SNVs with a heteroplasmy level of >5%. Variants were filtered using the following criteria: at poly-C homopolymer regions, read depth <300, and or with base quality, strandedness, slippage, weak evidence, germline, or position flags.…”