Deciphering Novel Chromosomal Structural Variants by Low-pass Whole Genome Sequencing in the Human Genome

Chromosomal microarray analysis (CMA) is the first-tier genetic test for chromosomal disorders and copy-number variants (CNVs) in prenatal and postnatal genetic diagnosis. Recently next-generation sequencing (NGS)-based CNV detection has emerged; however, it has not yet been well established and validated.Firstly, we conducted retrospective analyses of the spectrum and characteristics of clinically relevant CNVs in women undergoing prenatal testing by CMA to serve as a reference dataset to further develop low-pass (low-coverage and high throughput) genome sequencing (GS). Overall, clinically relevant CNVs (<10Mb in size) were detected in 1.7% (375/23,865) of pregnancies and about one-third of CNVs were less than 1Mb in size. Most occurred de novo; however, one-third were inherited from one of the parents. The findings have implications on prenatal screening and invasive diagnostic testing.Secondly, we aimed to evaluate the performance of low-pass GS for CNV detection. We investigated different sequencing parameters such as read length and read amount and their effects on the resolution and sensitivity of CNV detection, including mosaic CNVs. The results showed low-pass GS utilising 15 million single-end 50-bp reads (equivalent to 0.25X) were the optimal parameters in detecting genome-wide heterozygous CNVs at a resolution of 50-kb and 10-kb for homozygous or hemizygous CNV deletions. Low-pass GS also detected mosaic CNVs at a level as low as 30%.Thirdly, we conducted a retrospective comparison between low-pass GS and CMA in a cohort of 532 cases including prenatal, postnatal, and products of gestation samples to investigate its analytic utility. Low-pass GS detected all chromosomal imbalances and pathogenic CNVs previously ascertained by CMA, and reported four additional pathogenic or likely pathogenic findings, providing a relative improvement of diagnostic yield of ~3.4%. The additional findings comprised of small CNVs beyond the resolution of CMA and CNVs in regions that are not well covered by sufficient number of probes. To enlarge the sample size and evaluate the accuracy, efficacy, and incremental yield of low-pass GS for CNV detection, we further performed low-pass GS in parallel with CMA on a prospective clinical cohort of 1,023 singleton pregnancies. Low-pass GS demonstrated an equivalent ability in the detection of CNVs compared to CMA and provided additional diagnosis in 1.7% of cases. Low-pass GS provided higher sample throughput, a lower requirement of DNA amount, and a significantly lower experimental repeat rate due to QC failures. These two studies provided compelling evidence to make use of low-pass GS as a standard part of prenatal testing.Low-pass GS utilising large-insert size libraries (namely mate-pair genome sequencing) offered improved ability to detect novel structural rearrangements. Low-pass mate-pair GS was performed on 16 cases and revealed that most (68.6%) insertions involved additional and previously underappreciated cryptic complex rearrangements. The breakpoint characteristics and complex rearrangements suggested that chromosomal insertions were likely formed by DNA-replication-based mechanisms.Overall, we performed analytical and clinical validation of low-pass GS for the detection of chromosomal imbalances across various applications, identified novel structural rearrangements and their relationship with human diseases. Our studies demonstrated that low-pass GS can replace CMA and/karyotyping as a first-tier genetic tests for clinical cytogenetic investigation.染色體微陣列分析(CMA)目前為產前和兒科遺傳診斷拷貝數變異(CNV)的首選檢測手段。最近，二代測序展示了其檢測CNV的能力，但尚未進行大規模驗證。首先，我們對產前診斷樣本中CMA檢測出的具有臨床意義的CNV進行分析，研究其分佈和特徵，作為低深度高通量基因組測序(Low-pass GS)發展的参考基础。在接受CMA的孕婦中, 1.7%（375/23,865）檢測出具有臨床意義的CNV（