State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing 210096, China.
Jiangsu Cancer Hospital, Jiangsu Institute of Cancer Research, Nanjing Medical University Affiliated Cancer Hospital, Nanjing 210009, China.
Gene. 2019 May 30;699:145-154. doi: 10.1016/j.gene.2019.03.016. Epub 2019 Mar 13.
Detecting heteroplasmic variations in the mitochondrial genome can help identify potential pathogenic possibilities, which is significant for disease prevention. The development of next-generation sequencing changed the quantification of mitochondrial DNA (mtDNA) heteroplasmy from scanning limited recorded points to the entire mitochondrial genome. However, due to the presence of nuclear mtDNA homologous sequences (nuMTs), maximally retaining real variations while excluding falsest heteroplasmic variations from nuMTs and sequencing errors presents a dilemma.
Herein, we used an improved method for detecting low-frequency mtDNA heteroplasmic variations from whole genome sequencing data, including point variations and short-fragment length alterations, and evaluated the effect of this method. A two-step alignment was designed and performed to accelerate data processing, to obtain and retain the true mtDNA reads and to eliminate most nuMTs reads. After analyzing whole genome sequencing data of K562 and GM12878 cells, ~90% of heteroplasmic point variations were identified in MitoMap. The results were consistent with the results of an amplification refractory mutation system qPCR. Many linkages of the detected heteroplasmy variations were also discovered.
Our improved method is a simple, efficient and accurate way to mine mitochondrial low-frequency heteroplasmic variations from whole genome sequencing data. By evaluating the highest misalignment possibility caused by the remaining nuMTs-like reads and sequencing errors, our procedure can detect mtDNA heteroplasmic variations whose heteroplasmy frequencies are as low as 0.2%.
检测线粒体基因组中的异质体变异有助于识别潜在的致病可能性，这对于疾病预防具有重要意义。下一代测序技术的发展改变了从有限记录点扫描到整个线粒体基因组的线粒体 DNA（mtDNA）异质体定量方法。然而，由于存在核线粒体同源序列（nuMTs），在最大限度地保留真实变异的同时，排除来自 nuMTs 和测序错误的假异质体变异，这是一个难题。
本文介绍了一种从全基因组测序数据中检测低频 mtDNA 异质体变异的改进方法，包括点变异和短片段长度改变，并评估了该方法的效果。设计并执行了两步比对，以加速数据处理，获取并保留真实的 mtDNA 读数，并消除大多数 nuMTs 读数。在分析 K562 和 GM12878 细胞的全基因组测序数据后，约 90%的异质体点变异在 MitoMap 中被识别。结果与扩增阻滞突变系统 qPCR 的结果一致。还发现了许多检测到的异质体变异的关联。
我们的改进方法是一种从全基因组测序数据中挖掘线粒体低频异质体变异的简单、高效、准确的方法。通过评估由残留的类似 nuMTs 的读数和测序错误引起的最高错配可能性，我们的程序可以检测异质体频率低至 0.2%的 mtDNA 异质体变异。
