Chastity of the captured sequence data: insight from mitochondrial genome sequencing

With the development of next-generation sequencing (NGS) technologies, whole genome sequencing has become feasible for most organisms, nevertheless, targeted sequencing is still providing a more efficient way for most projects, as it takes better advantage of the sequencing capacity. PCR-based method has long been regarding as the gold standard for target-enrichment due to its higher specificity and reproducibility. However, recently, as more targets and samples are included in the experiment, capture method was proposed to fit better and is being widely used now, but there is a big concern on the homologous sequences which could be co-enriched in the sequencing library. In this study, existence of false-target contamination (nuclear mitochondrial DNA) was proven by examining the mitochondrial genome captured sequence data. Although their frequencies are normally lower than 5%, distinguishing them from the low-level mutation (heterplasmy) is still a big challenge, several possible solutions were discussed here, trying to optimize the balance between false positive and false negative. Therefore, special caution should be taken when processing the captured sequence data, as it does not only consist of off-target reads, but also could have false-target reads, proportion of the contamination may vary a lot for different targets.