Structural variation and fusion detection using targeted sequencing data from circulating cell free DNA
Structural variation and fusion detection using targeted sequencing data from circulating cell free DNA
复制标题
DOI：
10.1093/nar/gkz067
复制
发表时间：
2019-04-23
影响因子：
14.9
通讯作者：
Hach, Faraz
中科院分区：
文献类型：
--
作者：
Gawronskii, Alexander R.;Lin, Yen-Yi;Hach, Faraz
动机癌症是一种复杂的疾病，涉及快速进化的细胞，通常形成多个不同的克隆。为了有效了解患者特异性肿瘤的进展，需要在多个时间点对肿瘤 DNA 进行全面采样，最好通过廉价且微创的技术获得。当前的测序技术使液体活检成为可能，其中包括对患者的血液或尿液进行取样并对循环游离 DNA (cfDNA) 进行测序。该 DNA 的一定比例源自肿瘤，称为循环肿瘤 DNA (ctDNA)。样本中ctDNA的比例可能极低，并且ctDNA可能源自多个肿瘤或克隆。这些因素对将现有工具和工作流程应用于 ctDNA 分析提出了独特的挑战，特别是在结构变异检测方面，这依赖于足够的读段覆盖率才能检测到。结果在这里，我们介绍 SViCT，一种结构变异 (SV) 检测工具，旨在应对与 cfDNA 分析相关的挑战。 SViCT 可以检测断点和各种结构变异的序列，包括删除、插入、倒位、重复和易位。 SViCTextracts 不一致的读段对、一端锚定和软剪辑/分割读段，将它们组装成重叠群，并使用有效的 k-mer 索引方法将重叠群间隔重新映射到参考基因组。然后使用图和贪婪算法的组合来连接这些间隔，以识别特定的结构变体签名。我们评估了 SViCT 的性能，并将其与使用模拟 cfDNA 数据集（其属性与真实 cfDNA 样本的属性相匹配）的最先进工具进行了比较。我们的工具的阳性预测值和灵敏度优于所有测试的工具，并且在模拟数据集中低至 0.01% 肿瘤 DNA 的最低稀释度下仍保持合理的性能。此外，SViCT 能够检测两个真实 cfDNA 参考数据集中（0.6-5% ctDNA）中的所有已知 SV，并预测前列腺癌队列中的新结构变异。可用性：SViCT 可在 https://github.com/vpc-ccg/svict 上获取。联系方式：faraz.hach@ubc.ca
Motivation Cancer is a complex disease that involves rapidly evolving cells, often forming multiple distinct clones. In order to effectively understand progression of a patient-specific tumor, one needs to comprehensively sample tumor DNA at multiple time points, ideally obtained through inexpensive and minimally invasive techniques. Current sequencing technologies make the liquid biopsy' possible, which involves sampling a patient's blood or urine and sequencing the circulating cell free DNA (cfDNA). A certain percentage of this DNA originates from the tumor, known as circulating tumor DNA (ctDNA). The ratio of ctDNA may be extremely low in the sample, and the ctDNA may originate from multiple tumors or clones. These factors present unique challenges for applying existing tools and workflows to the analysis of ctDNA, especially in the detection of structural variations which rely on sufficient read coverage to be detectable.Results Here we introduce SViCT, a structural variation (SV) detection tool designed to handle the challenges associated with cfDNA analysis. SViCTcan detect breakpoints and sequences of various structural variations including deletions, insertions, inversions, duplications and translocations. SViCTextracts discordant read pairs, one-end anchors and soft-clipped/split reads, assembles them into contigs, and re-maps contig intervals to a reference genome using an efficient k-mer indexing approach. The intervals are then joined using a combination of graph and greedy algorithms to identify specific structural variant signatures. We assessed the performance of SViCTand compared it to state-of-the-art tools using simulated cfDNA datasets with properties matching those of real cfDNA samples. The positive predictive value and sensitivity of our tool was superior to all the tested tools and reasonable performance was maintained down to the lowest dilution of 0.01% tumor DNA in simulated datasets. Additionally, SViCTwas able to detect all known SVs in two real cfDNA reference datasets (at 0.6-5% ctDNA) and predict a novel structural variant in a prostate cancer cohort.Availability: SViCT is available at https://github.com/vpc-ccg/svict. Contact: faraz.hach@ubc.ca