Whole-genome sequencing (WGS) of viruses from patient or environmental samples can provide tremendous insight into the epidemiology, drug resistance or evolution of a virus. However, we face two common hurdles in obtaining robust sequence information; the low copy number of viral genomes in specimens and the error introduced by WGS techniques. To optimize detection and minimize error in WGS of hantaviruses, we tested four amplification approaches and different amplicon pooling methods for library preparation and examined these preparations using two sequencing platforms, Illumina MiSeq and Oxford Nanopore Technologies MinION. First, we tested and optimized primers used for whole segment PCR or one kilobase amplicon amplification for even coverage using RNA isolated from the supernatant of virus-infected cells. Once optimized we assessed two sources of total RNA, virus-infected cells and supernatant from the virus-infected cells, with four variations of primer pooling for amplicons, and six different amplification approaches. We show that 99–100% genome coverage was obtained using a one-step RT-PCR reaction with one forward and reverse primer. Using a two-step RT-PCR with three distinct tiling approaches for the three genomic segments (vRNAs), we optimized primer pooling approaches for PCR amplification to achieve a greater number of aligned reads, average depth of genome, and genome coverage. The single nucleotide polymorphisms identified from MiSeq and MinION sequencing suggested intrinsic mutation frequencies of ~10⁻⁵-10⁻⁷ per genome and 10⁻⁴-10⁻⁵ per genome, respectively. We noted no difference in the coverage or accuracy when comparing WGS results with amplicons amplified from RNA extracted from infected cells or supernatant of these infected cells. Our results show that high-throughput diagnostics requiring the identification of hantavirus species or strains can be performed using MiSeq or MinION using a one-step approach. However, the two-step MiSeq approach outperformed the MinION in coverage depth and accuracy, and hence would be superior for assessment of genomes for epidemiology or evolutionary questions using the methods developed herein.

对患者或环境样本中的病毒进行全基因组测序 (WGS) 可以深入了解病毒的流行病学、耐药性或进化。然而，我们在获得可靠的序列信息方面面临两个常见的障碍：样本中病毒基因组的低拷贝数以及全基因组测序技术引入的误差。为了优化汉坦病毒 WGS 的检测并最大程度地减少错误，我们测试了四种用于文库制备的扩增方法和不同的扩增子汇集方法，并使用 Illumina MiSeq 和 Oxford Nanopore Technologies MinION 这两个测序平台检查了这些制备物。首先，我们使用从病毒感染细胞上清液中分离的 RNA 测试并优化了用于全片段 PCR 或 1 KB 扩增子扩增的引物，以实现均匀覆盖。优化后，我们评估了两种来源的总 RNA、病毒感染的细胞和病毒感染细胞的上清液，以及扩增子引物池的四种变体和六种不同的扩增方法。我们证明，使用一种正向和反向引物的一步 RT-PCR 反应可以获得 99-100% 的基因组覆盖率。我们使用两步 RT-PCR 以及针对三个基因组片段 (vRNA) 的三种不同平铺方法，优化了 PCR 扩增的引物池方法，以实现更多的比对读数、基因组平均深度和基因组覆盖率。从 MiSeq 和 MinION 测序中鉴定出的单核苷酸多态性表明，每个基因组的固有突变频率分别约为 10 ^-5 -10 ^-7和每个基因组 10 ^-4 -10 ^-5 。 当将 WGS 结果与从感染细胞或这些感染细胞上清液中提取的 RNA 扩增的扩增子进行比较时，我们注意到覆盖率或准确性没有差异。我们的结果表明，需要鉴定汉坦病毒种类或毒株的高通量诊断可以使用 MiSeq 或 MinION 使用一步法进行。然而，两步 MiSeq 方法在覆盖深度和准确性方面优于 MinION，因此使用本文开发的方法评估流行病学或进化问题的基因组会更有效。