Direct prediction of carbapenem-resistance for Pseudomonas aeruginosa by whole genome sequencing and metagenomic sequencing

Carbapenem-resistance is a major concern in the management of antibiotic resistant Pseudomonas aeruginosa infections. Direct prediction of carbapenem-resistance phenotype from genotype in P.aeruginosa isolates and clinical samples would promote timely antibiotic therapy. While the complex carbapenem-resistance mechanism and the high prevalence of variant-driven carbapenem-resistance in P.aeruginosa make it challenging to predict its carbapenem-resistance phenotype through genotype. In this study, by use of whole genome sequencing (WGS) data of 1622 P.aeruginosa isolates followed by machine learning, we screened out 16 and 31 key gene features associated with carbapenem-resistance of P.aeruginosa, including oprD(HIGH), and constructed the resistance prediction models regarding imipenem (IPM) or meropenem (MEM). The Area Under Curves of IPM- and MEM-resistance prediction models were 0.906 and 0.925. For the direct carbapenem-resistance prediction for P. aeruginosa from clinical samples by those formerly selected key gene features and constructed prediction models, 72 P. aeruginosa-positive sputum samples were collected and sequenced by metagenomic sequencing (MGS) based on next-generation sequencing (NGS) or Oxford Nanopore Technology (ONT). It turned out that the prediction applicability of MGS based on NGS outperformed that of MGS based on ONT. In 72 P. aeruginosa-positive sputum samples, 65.0% (26/40) IPM-insensitive and 63.2% (24/38) MEM-insensitive P. aeruginosa could be directly predicted by NGS-based MGS with PPVs of 0.897 and 0.889. By the direct detection of those selected key gene features associated with carbapenem-resistance of P. aeruginosa, carbapenem-resistance of P. aeruginosa could be directly predicted from cultured isolates by WGS or clinical samples by NGS-based MGS, which could assist the timely treatment and surveillance of carbapenem-resistant P.aeruginosa.

碳青霉烯类耐药（Carbapenem-resistance）是耐药铜绿假单胞菌（Pseudomonas aeruginosa）感染临床管理中的重大关切问题。直接从铜绿假单胞菌分离株与临床样本的基因型预测其碳青霉烯类耐药表型，有助于及时开展针对性抗生素治疗。然而，铜绿假单胞菌复杂的碳青霉烯类耐药机制，以及变异驱动的碳青霉烯类耐药高流行率，使得通过基因型预测其耐药表型颇具挑战。本研究针对1622株铜绿假单胞菌分离株的全基因组测序（whole genome sequencing, WGS）数据开展机器学习分析，分别筛选出16个和31个与铜绿假单胞菌碳青霉烯类耐药相关的关键基因特征（包括oprD(HIGH)），并据此构建了针对亚胺培南（imipenem, IPM）和美罗培南（meropenem, MEM）的耐药预测模型。亚胺培南与美罗培南耐药预测模型的曲线下面积（Area Under Curves, AUC）分别为0.906与0.925。为利用上述筛选得到的关键基因特征与构建好的预测模型，直接从临床样本中预测铜绿假单胞菌的碳青霉烯类耐药性，本研究收集了72份铜绿假单胞菌阳性痰液样本，并分别采用二代测序（next-generation sequencing, NGS）与牛津纳米孔技术（Oxford Nanopore Technology, ONT）完成宏基因组测序（metagenomic sequencing, MGS）。结果显示，基于二代测序的宏基因组测序其预测应用性能优于基于牛津纳米孔技术的宏基因组测序。在72份铜绿假单胞菌阳性痰液样本中，基于二代测序的宏基因组测序可直接预测出65.0%（26/40）的亚胺培南不敏感株与63.2%（24/38）的美罗培南不敏感株，其阳性预测值（Positive Predictive Value, PPV）分别为0.897与0.889。通过直接检测上述筛选得到的铜绿假单胞菌碳青霉烯类耐药相关关键基因特征，即可通过全基因组测序对培养分离株进行碳青霉烯类耐药性预测，或通过基于二代测序的宏基因组测序对临床样本进行耐药性预测，从而为碳青霉烯类耐药铜绿假单胞菌的及时治疗与监测提供辅助支持。