[Application of Machine Learning Techniques for Antimicrobial Resistance Prediction].

Abstract

Due to the abuse or excessive use of antimicrobials， particularly antibiotics， antimicrobial resistance （AMR） has become one of the major challenges in global public health. The rapid growth of microbial data， facilitated by advancements in high-throughput sequencing technology， underscores the importance of leveraging machine learning for predicting AMR and identifying resistance markers. Machine learning， encompassing supervised and unsupervised learning， has been proven effective by early studies of AMR prediction. By analyzing microbial genomes and AMR data to build machine learning models， we can improve predictions of microbial resistance and develop more effective antibiotic use strategies， thereby controlling the spread of resistance. This review article focuses on the specific construction processes of machine learning algorithms and the models commonly employed in AMR studies. It also highlights the diverse applications and prospects of machine learning in AMR prediction， with the goal of offering a scientific foundation for future environmental AMR monitoring initiatives.