Scholarlink

Antimicrobial resistance (AMR) presents a significant global challenge impacting human and animal health, particularly concerning pathogens such as Chlamydiae species. Traditional methods for antimicrobial susceptibility testing can be time-consuming and heavily reliant on bacterial cultures. To tackle these issues, our study employs Machine Learning (ML) to automate AMR prediction based on bacterial protein sequences. We have developed a specialized ML tool to predict whether specific protein sequences of Chlamydiae strains exhibit resistance or sensitivity. This tool encompasses essential processes, including data collection, preprocessing, data encoding, model selection, training, and evaluation. Among the models assessed, the Gradient Boosting Machine (GBM) demonstrated the highest overall performance, achieving an accuracy of 0.83, a precision of 0.87, an F1-score of 0.82, and an AUROC of 0.83, indicating excellent discriminatory capability. To enhance accessibility, we have deployed this model within a user-friendly web application built using the Streamlit framework. This application enables researchers and healthcare professionals to efficiently and rapidly predict AMR. This tool significantly advances ML for rapid and reliable predictions of antimicrobial resistance in Chlamydiae protein sequences, addressing a crucial need in combating antibiotic resistance in these pathogens.

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

Machine Learning-Based Prediction of Antimicrobial Resistance Using Chlamydiae Protein Sequences

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sanae Esskhayry,
  • Ichrak Benamri,
  • Afaf Lamzouri,
  • Ouafae Kaissi,
  • Rachida Fissoune,
  • Ahmed Moussa,
  • Fouzia Radouani

摘要

Antimicrobial resistance (AMR) presents a significant global challenge impacting human and animal health, particularly concerning pathogens such as Chlamydiae species. Traditional methods for antimicrobial susceptibility testing can be time-consuming and heavily reliant on bacterial cultures. To tackle these issues, our study employs Machine Learning (ML) to automate AMR prediction based on bacterial protein sequences. We have developed a specialized ML tool to predict whether specific protein sequences of Chlamydiae strains exhibit resistance or sensitivity. This tool encompasses essential processes, including data collection, preprocessing, data encoding, model selection, training, and evaluation. Among the models assessed, the Gradient Boosting Machine (GBM) demonstrated the highest overall performance, achieving an accuracy of 0.83, a precision of 0.87, an F1-score of 0.82, and an AUROC of 0.83, indicating excellent discriminatory capability. To enhance accessibility, we have deployed this model within a user-friendly web application built using the Streamlit framework. This application enables researchers and healthcare professionals to efficiently and rapidly predict AMR. This tool significantly advances ML for rapid and reliable predictions of antimicrobial resistance in Chlamydiae protein sequences, addressing a crucial need in combating antibiotic resistance in these pathogens.