Center for Personalized Medicine, Department of Pathology and Laboratory Medicine, Children's Hospital Los Angeles, Los Angeles, California; Keck School of Medicine of University of Southern California, Los Angeles, California.
Center for Personalized Medicine, Department of Pathology and Laboratory Medicine, Children's Hospital Los Angeles, Los Angeles, California.
J Mol Diagn. 2024 Feb;26(2):127-139. doi: 10.1016/j.jmoldx.2023.11.003. Epub 2023 Nov 24.
This study reports the development of an exome capture-based RNA-sequencing assay to detect recurring and novel fusions in hematologic, solid, and central nervous system tumors. The assay used Twist Comprehensive Exome capture with either fresh or formalin-fixed samples and a bioinformatic platform that provides fusion detection, prioritization, and downstream curation. A minimum of 50 million uniquely mapped reads, a consensus read alignment/fusion calling approach using four callers (Arriba, FusionCatcher, STAR-Fusion, and Dragen), and custom software were used to integrate, annotate, and rank the candidate fusion calls. In an evaluation of 50 samples, the number of calls varied substantially by caller, from a mean of 24.8 with STAR-Fusion to 259.6 with FusionCatcher; only 1.1% of calls were made by all four callers. Therefore a filtering and ranking algorithm was developed based on multiple criteria, including number of supporting reads, calling consensus, genes involved, and cross-reference against databases of known cancer-associated or likely false-positive fusions. This approach was highly effective in pinpointing known clinically relevant fusions, ranking them first in 47 of 50 samples (94%). Detection of pathogenic gene fusions in three diagnostically challenging cases highlights the importance of a genome-wide and nontargeted method for fusion detection in pediatric cancer.
本研究报告了一种基于外显子组捕获的 RNA 测序检测方法的开发，用于检测血液、实体和中枢神经系统肿瘤中的复发和新型融合。该检测方法使用 Twist Comprehensive Exome 捕获试剂盒，结合新鲜或福尔马林固定样本，并使用一个提供融合检测、优先级排序和下游整理的生物信息学平台。使用最少 5000 万条唯一映射的读段、一种使用四个调用者（Arriba、FusionCatcher、STAR-Fusion 和 Dragen）的共识读对齐/融合调用方法，以及定制软件来整合、注释和对候选融合调用进行排名。在对 50 个样本的评估中，调用者之间的调用数量差异很大，从 STAR-Fusion 的平均 24.8 个到 FusionCatcher 的 259.6 个；只有 1.1%的调用是由所有四个调用者做出的。因此，开发了一种基于多种标准的过滤和排名算法，包括支持读段的数量、调用共识、涉及的基因以及与已知癌症相关或可能的假阳性融合的数据库的交叉引用。这种方法在确定已知的临床相关融合方面非常有效，在 50 个样本中的 47 个（94%）样本中首先对其进行了排名。在三个诊断具有挑战性的病例中检测到致病性基因融合，突出了在儿科癌症中进行融合检测的全基因组和非靶向方法的重要性。
