Scholarlink

<p>Brucellosis is a highly infectious anthropozoonotic disease caused by the genus Brucella, which is the pathogenic bacteria in public health risk and occupation exposure. Therefore, rapid detection and identification of Brucella is crucial for controlling and managing the prevalence of brucellosis. In this work, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) combined with machine learning (ML) algorithms was developed to identify Brucella isolates at the species level rapidly. A total of 1440 MALDI spectral data from 59 strains of <i>B. melitensis</i> (23 strains of <i>B. melitensis</i> bv. 1 and 36 strains of <i>B. melitensis</i> bv. 3) were obtained. Four ML algorithms including support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), and logistic regression (LR) were compared for typing Brucella species. The area under receiver operating characteristic (ROC) curve, precision-recall ratio (PR), accuracy, specificity, and sensitivity were used to evaluate the performances of different ML algorithms. These results showed that MALDI-TOF MS combined with the XGBoost algorithm is the best method for typing Brucella species with good performances, demonstrating the significant potential applications in the diagnosis, control, and epidemiology of brucellosis.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Typing Brucella Species by MALDI-TOF Mass Spectrometry Combined with Machine Learning

  • 期刊论文

  • 发表日期：

    2025年2月28日

查看全文
  • Hua Cai,
  • Guizhen Wang,
  • Yuzhen Bai,
  • Haitao Yuan,
  • Liping Feng,
  • Hong Wang,
  • Changshan Guo,
  • Huitian Li,
  • Lina Liu,
  • Yunxia Pu,
  • Jiawei Shi,
  • Ping Liu,
  • Shibo Wang,
  • Dong Zhang,
  • Fangang Zeng,
  • Bin Hu

摘要

Brucellosis is a highly infectious anthropozoonotic disease caused by the genus Brucella, which is the pathogenic bacteria in public health risk and occupation exposure. Therefore, rapid detection and identification of Brucella is crucial for controlling and managing the prevalence of brucellosis. In this work, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) combined with machine learning (ML) algorithms was developed to identify Brucella isolates at the species level rapidly. A total of 1440 MALDI spectral data from 59 strains of B. melitensis (23 strains of B. melitensis bv. 1 and 36 strains of B. melitensis bv. 3) were obtained. Four ML algorithms including support vector machine (SVM), random forest (RF), extreme gradient boosting (XGBoost), and logistic regression (LR) were compared for typing Brucella species. The area under receiver operating characteristic (ROC) curve, precision-recall ratio (PR), accuracy, specificity, and sensitivity were used to evaluate the performances of different ML algorithms. These results showed that MALDI-TOF MS combined with the XGBoost algorithm is the best method for typing Brucella species with good performances, demonstrating the significant potential applications in the diagnosis, control, and epidemiology of brucellosis.