Background To date, the usage of Galaxy, an open-source bioinformatics platform, has been reported primarily in research. We report 5 years’ experience (2015 to 2020) with Galaxy in our hospital, as part of the “Assistance Publique–Hôpitaux de Paris” (AP-HP), to demonstrate its suitability for high-throughput sequencing (HTS) data analysis in a clinical laboratory setting. Methods Our Galaxy instance has been running since July 2015 and is used daily to study inherited diseases, cancer, and microbiology. For the molecular diagnosis of hereditary diseases, 6970 patients were analyzed with Galaxy (corresponding to a total of 7029 analyses). Results Using Galaxy, the time to process a batch of 23 samples—equivalent to a targeted DNA sequencing MiSeq run—from raw data to an annotated variant call file was generally less than 2 h for panels between 1 and 500 kb. Over 5 years, we only restarted the server twice for hardware maintenance and did not experience any significant troubles, demonstrating the robustness of our Galaxy installation in conjunction with HTCondor as a job scheduler and a PostgreSQL database. The quality of our targeted exome sequencing method was externally evaluated annually by the European Molecular Genetics Quality Network (EMQN). Sensitivity was mean (SD)% 99 (2)% for single nucleotide variants and 93 (9)% for small insertion-deletions. Conclusion Our experience with Galaxy demonstrates it to be a suitable platform for HTS data analysis with vast potential to benefit patient care in a clinical laboratory setting.

中文翻译：

---

Galaxy 是一个适用于使用高通量测序数据分析对人类遗传疾病进行分子诊断的生物信息学平台：临床实验室的五年经验

背景 迄今为止，主要在研究中报道了开源生物信息学平台 Galaxy 的使用。我们报告了 Galaxy 在我们医院 5 年（2015 年至 2020 年）的经验，作为“Assistance Publique–Hôpitaux de Paris” (AP-HP) 的一部分，以证明其适用于高通量测序 (HTS) 数据分析临床实验室环境。方法 我们的 Galaxy 实例自 2015 年 7 月开始运行，每天用于研究遗传性疾病、癌症和微生物学。对于遗传性疾病的分子诊断，使用 Galaxy 分析了 6970 名患者（对应总共 7029 次分析）。使用 Galaxy 的结果，对于 1 到 500 kb 的面板，处理一批 23 个样本的时间（相当于靶向 DNA 测序 MiSeq 运行）从原始数据到带注释的变异调用文件通常不到 2 小时。5 年多来，我们只为硬件维护重新启动了两次服务器，并且没有遇到任何重大问题，这证明了我们的 Galaxy 安装与 HTCondor 作为作业调度程序和 PostgreSQL 数据库的稳健性。我们的靶向外显子组测序方法的质量每年由欧洲分子遗传学质量网络 (EMQN) 进行外部评估。单核苷酸变体的灵敏度为平均 (SD)% 99 (2)%，小插入缺失的灵敏度为 93 (9)%。