Clinical long-read sequencing of the human mitochondrial genome for mitochondrial disease diagnostics

Purpose: Long-read, third generation, sequencing technologies have the potential to improve current state of the art diagnostic strategies. In order to determine if long-read sequencing technologies are suitable for diagnosis of mitochondrial disorders due to mitochondrial DNA (mtDNA) variants, particularly large deletions, we compared the performance of Oxford Nanopore Technologies (ONT) MinION to current diagnostic methods. Methods: We sequenced mtDNA from nine patients with mtDNA deletion disorders and three controls with both ONT MinION and Illumina MiSeq. We applied a computational pipeline to estimate the positions of mtDNA deletions in patients, and subsequently validated the breakpoints using Sanger sequencing. Results: We were able to detect mtDNA deletions with a MinION workflow, successfully calling the disease causing event in all cases. Sequencing coverage was in most cases significantly more (p=0.03, Wilcoxon test) uniform with MinION than with MiSeq and subsequent correction of MinION reads improved breakpoint accuracy and reduced false positives. Although heteroplasmic single nucleotide variants are detectable, the high number of false positives and false negatives precludes their use in diagnostics at this time. Conclusion: The MinION is becoming an increasingly attractive diagnostic tool due to the reducing cost, increasing accuracy, and the speed at which data can be obtained.