Howell Kate J, Weinert Lucy A, Peters Sarah E, Wang Jinhong, Hernandez-Garcia Juan, Chaudhuri Roy R, Luan Shi-Lu, Angen Øystein, Aragon Virginia, Williamson Susanna M, Langford Paul R, Rycroft Andrew N, Wren Brendan W, Maskell Duncan J, Tucker Alexander W
Department of Veterinary Medicine, University of Cambridge, Cambridge, United Kingdom
Department of Veterinary Medicine, University of Cambridge, Cambridge, United Kingdom.
J Clin Microbiol. 2017 Sep;55(9):2617-2628. doi: 10.1128/JCM.02464-16. Epub 2017 Jun 14.
is a diverse bacterial species that is found in the upper respiratory tracts of pigs and can also cause Glässer's disease and pneumonia. A previous pangenome study of identified 48 genes that were associated with clinical disease. Here, we describe the development of a generalized linear model (termed a pathotyping model) to predict the potential virulence of isolates of based on a subset of 10 genes from the pangenome. A multiplex PCR (mPCR) was constructed based on these genes, the results of which were entered into the pathotyping model to yield a prediction of virulence. This new diagnostic mPCR was tested on 143 field isolates of that had previously been whole-genome sequenced and a further 84 isolates from the United Kingdom from cases of related disease in pigs collected between 2013 and 2014. The combination of the mPCR and the pathotyping model predicted the virulence of an isolate with 78% accuracy for the original isolate collection and 90% for the additional isolate collection, providing an overall accuracy of 83% (81% sensitivity and 93% specificity) compared with that of the "current standard" of detailed clinical metadata. This new pathotyping assay has the potential to aid surveillance and disease control in addition to serotyping data.
是一种多样的细菌物种,存在于猪的上呼吸道中,也可引起格拉泽氏病和肺炎。先前对该细菌的泛基因组研究鉴定出48个与临床疾病相关的基因。在此,我们描述了一种广义线性模型(称为致病型分型模型)的开发,该模型基于泛基因组中的10个基因子集来预测该细菌分离株的潜在毒力。基于这些基因构建了多重PCR(mPCR),其结果输入致病型分型模型以得出毒力预测。这种新的诊断性mPCR在143株先前已进行全基因组测序的该细菌田间分离株以及另外84株于2013年至2014年期间从英国猪的相关疾病病例中收集的分离株上进行了测试。mPCR与致病型分型模型的组合对原始分离株集合预测分离株毒力的准确率为78%,对额外分离株集合的准确率为90%,与详细临床元数据的“当前标准”相比,总体准确率为83%(敏感性81%,特异性93%)。除血清分型数据外,这种新的致病型分型检测方法有潜力辅助监测和疾病控制。