Tzou Philip L, Rhee Soo-Yon, Shafer Robert W
Division of Infectious Diseases, Department of Medicine, Stanford University, Stanford, California.
AIDS Res Hum Retroviruses. 2019 Oct;35(10):924-929. doi: 10.1089/AID.2018.0289. Epub 2019 Aug 26.
The presence of many highly unusual HIV-1 mutations at a minority variant threshold by next-generation sequence (NGS) may indicate that a high proportion of variants at or just above the threshold represent PCR errors. The validity of this hypothesis depends on the concept that highly unusual mutations detected by population-based sequencing are also highly unusual within a person's virus population. Highly unusual mutations were defined as mutations with a prevalence <0.01% in group M HIV-1 direct PCR population-based Sanger sequences in the Stanford HIV Drug Resistance Database. Single genome Sanger sequences [single genome sequences (SGSs)] were analyzed because they are not subject to PCR error. Permutation analyses compared the proportion of highly unusual mutations in SGSs with the empirical frequencies of these mutations in repeated random selections of population-based sequences. We created a database of 11,258 SGSs in 963 plasma samples from 345 persons with active virus replication and analyzed the subset of samples containing 10 or more SGSs. Highly unusual mutations occurred more commonly in samples undergoing SGS compared with population-based sequencing in protease (3.9% vs. 0.8%; < .001), reverse transcriptase (6.5% vs. 1.5%; < .001), and integrase (5.0% vs. 1.8%; < .001). Highly unusual mutations occur more commonly in SGSs than in population-based sequences. However, they comprise a small proportion of all SGS mutations supporting the concept that the presence of many highly unusual mutations just above an NGS threshold suggests that the threshold is too low.
通过下一代测序（NGS）在少数变异体阈值下出现许多高度异常的HIV-1突变，可能表明阈值处或略高于阈值的变异体中有很大一部分代表PCR错误。这一假设的有效性取决于这样一个概念，即基于群体测序检测到的高度异常突变在一个人的病毒群体中也是高度异常的。高度异常突变被定义为在斯坦福HIV耐药数据库中基于群体的M组HIV-1直接PCR Sanger序列中流行率<0.01%的突变。对单基因组Sanger序列[单基因组序列（SGS）]进行分析，因为它们不受PCR错误的影响。置换分析将SGS中高度异常突变的比例与基于群体的序列重复随机选择中这些突变的经验频率进行比较。我们创建了一个数据库，包含来自345名病毒活跃复制者的963份血浆样本中的11258条SGS，并分析了包含10条或更多SGS的样本子集。与基于群体的测序相比，在蛋白酶（3.9%对0.8%；<0.001）、逆转录酶（6.5%对1.5%；<0.001）和整合酶（5.0%对1.8%；<0.001）中，高度异常突变在进行SGS的样本中更常见。高度异常突变在SGS中比在基于群体的序列中更常见。然而，它们在所有SGS突变中所占比例很小，这支持了这样一个概念，即略高于NGS阈值出现许多高度异常突变表明该阈值过低。
