Data quality of whole genome bisulfite sequencing on Illumina platforms

The powerful HiSeq X sequencers with their patterned flowcell technology and fast turnaround times are instrumental for many large-scale genomic and epigenomic studies. However, assessment of DNA methylation by sodium bisulfite treatment results in sequencing libraries of low diversity, which may impact data quality and yield. In this report we assess the quality of WGBS data generated on the HiSeq X system in comparison with data generated on the HiSeq 2500 system and the newly released NovaSeq system. We report a systematic issue with low basecall quality scores assigned to guanines in the second read of WGBS when using certain Real Time Analysis (RTA) software versions on the HiSeq X sequencer, reminiscent of an issue that was previously reported with certain HiSeq 2500 software versions. However, with the HD.3.4.0 /RTA 2.7.7 software upgrade for the HiSeq X system, we observed an overall improved quality and yield of the WGBS data generated, which in turn empowers cost-effective and high quality DNA methylation studies.

搭载图案化流动池（patterned flowcell）技术且周转效率优异的高性能HiSeq X测序仪，是众多大规模基因组学与表观基因组学研究的核心支撑工具。然而，利用亚硫酸氢盐处理开展DNA甲基化检测时，会得到多样性偏低的测序文库，进而可能对数据质量与测序产出造成负面影响。本研究针对HiSeq X测序平台、HiSeq 2500测序平台与最新发布的NovaSeq测序平台所产生的全基因组亚硫酸氢盐测序（Whole Genome Bisulfite Sequencing，WGBS）数据质量展开对比评估。研究发现，在HiSeq X测序仪上使用特定版本的实时分析（Real Time Analysis，RTA）软件时，WGBS测序的第二读段中的鸟嘌呤碱基会被赋予较低的碱基识别质量分数，这一系统性问题与此前部分HiSeq 2500软件版本所报道的问题高度相似。但针对HiSeq X平台的HD.3.4.0/RTA 2.7.7软件升级完成后，我们发现所产出的WGBS数据整体质量与测序产出均得到显著改善，由此可为兼具成本效益与高质量标准的DNA甲基化研究提供支撑。