Email Alert | RSS    帮助

中国防痨杂志 ›› 2020, Vol. 42 ›› Issue (11): 1196-1202.doi: 10.3969/j.issn.1000-6621.2020.11.010

• 论著 • 上一篇    下一篇

全基因组数据分析工具TB Profiler v2.8.0、Mykrobe v0.7.0和PhyResSE v1.0在耐药结核病检测中的价值

李冰莹, 郑旭彬, 胡屹, 徐飚   

  1. 200032 上海,复旦大学公共卫生学院流行病学教研室 国家卫生健康委员会卫生技术评估重点实验室
  • 收稿日期:2020-06-30 出版日期:2020-11-10 发布日期:2020-11-13
  • 基金资助:
    “十三五”科技部传染病重大专项(2018ZX-10715-012);国家自然科学基金面上项目(81773491)

Evaluation of the application value of the whole genome sequence analysis tools, TB Profiler v2.8.0, Mykrobe v0.7.0 and PhyResSE v1.0, in testing drug-resistant tuberculosis

LI Bing-ying, ZHENG Xu-bin, HU Yi, XU Biao   

  1. Department of Epidemiology, School of Public Health, Key Lab of Health Technology Assessment, National Health Commission, Fudan University, Shanghai 200032, China
  • Received:2020-06-30 Online:2020-11-10 Published:2020-11-13

摘要:

目的 评估3种针对结核分枝杆菌开发的全基因组数据分析工具,即TB Profiler v2.8.0、Mykrobe v0.7.0和PhyResSE v1.0(简称“TB Profiler、Mykrobe和PhyResSE”)在耐药结核病诊断中的性能。方法 从美国国立生物技术信息中心核酸数据库(National Center for Biotechnology Information Sequence Read Archive, NCBI SRA)收集了先前2项研究所上传的534株中国结核分枝杆菌临床分离株的全基因组测序数据和表型药物敏感性试验(drug susceptibility testing, DST)结果,其中包括457株耐多药菌株和77株敏感菌株。使用TB Profiler、Mykrobe和PhyResSE对全基因组数据进行分析,检测一线和二线抗结核药品的耐药性,并将其与DST结果进行比较,评价这3种工具的检测效能。结果 以DST结果为参照标准,TB Profiler、Mykrobe和PhyResSE检测利福平耐药的敏感度相近,分别为90.81%(415/457)、87.75%(401/457)和90.81%(415/457)。Mykrobe和PhyResSE检测异烟肼耐药的敏感度分别为76.42%(350/458)和76.20%(349/458),略高于TB Profiler(69.43%,318/458)。3种工具检测乙胺丁醇和链霉素耐药的敏感度相近,范围从76.00%到81.61%不等。对于吡嗪酰胺,TB Profiler的敏感度(72.82%,150/206)高于Mykrobe(61.65%,127/206)和PhyResSE(50.97%,105/206)。PhyResSE检测氟喹诺酮类和阿米卡星耐药的敏感度分别为88.27%(143/162)和60.00%(27/45),高于TB Profiler的81.48%(132/162)和48.89%(22/45)和Mykrobe的82.10%(133/162)和55.56%(25/45)。3种工具检测抗结核药品耐药的特异度相近并且均较高,除乙胺丁醇为82.42%~83.88%外,对其余药品的特异度均高于90%。结论 3种工具检测效能良好,可以快速检测抗结核药品耐药性,有良好的发展前景,但是目前对吡嗪酰胺和一些二线抗结核药品耐药检测的敏感度较低,需要加强耐药机制研究。

关键词: 分枝杆菌,结核, 全基因组测序, 诊断, 抗药性

Abstract:

Objective To evaluate the application of three whole genome sequence (WGS) analysis tools developed for Mycobacterium tuberculosis (MTB), TB Profiler v2.8.0, Mykrobe v0.7.0 and PhyResSE v1.0, in testing drug-resistant of tuberculosis. Methods WGS data and drug susceptibility testing (DST) results of 534 MTB clinical isolates from two previous studies were collected from National Center for Biotechnology Information Sequence Read Archive (NCBI SRA) database. Of the 534 MTB clinical isolates, 457 were multi-resistant and 77 were susceptible. Using DST as reference, WGS data were analyzed by TB Profiler, Mykrobe and PhyResSE to access the performance on predicting resistance to first-line and second-line anti-tuberculosis drugs. Results Taking DST results as reference, the sensitivities of TB Profiler, Mykrobe and PhyResSE for rifampicin resistance were similar, which were 90.81% (415/457), 87.75% (401/457) and 90.81% (415/457), respectively. The sensitivities of Mykrobe and PhyResSE for isoniazid resistance were 76.42% (350/458) and 76.20% (349/458), slightly higher than that of TB Profiler (69.43%, 318/458). The sensitivities of the three tools for ethambutol and streptomycin resistance were similar, ranging from 76.00% to 81.61%. As to pyrazinamide, the sensitivity of TB Profiler (72.82%, 150/206) was higher than those of Mykrobe (61.65%, 127/206) and PhyResSE (50.97%, 105/206). For fluoroquinolones and amikacin, the sensitivities of PhyResSE were 88.27% (143/162) and 60.00% (27/45), higher than those of TB Profiler (81.48% (132/162) and 48.89% (22/45)) and Mykrobe (82.10% (133/162) and 55.56% (25/45). The specificities of the three tools for detecting drug resistance of anti-tuberculosis drugs were similar and high (higher than 90%), except for ethambutol which was 82.42%-83.88%. Conclusion All of the three bioinformatics tools have a good performance on rapid detection of drug-resistant tuberculosis with a promising prospect of future application. The relatively low sensitivity of the tools to pyrazinamide and some second-line anti-tuberculosis drugs suggests that studies on resistance mechanism of these drugs should be enhanced.

Key words: Mycobacterium tuberculosis, Genome-wide sequencing, Diagnosis, Drug resistance