Genomics-driven risk assessment of antimicrobial resistance: Current status, challenges, and future perspectives

Antimicrobial resistance has become a major global public health challenge. However, conventional antimicrobial resistance risk assessment has largely focused on phenotypes while overlooking the dynamic dissemination of resistance genes and their mobility. This review systematically examines how genomics is reshaping microbial antimicrobial resistance risk assessment, with particular emphasis on the integration of key dimensions including clinically important antimicrobials, resistance gene mobility, evidence of ARG transmission along exposure pathways, and data quality and uncertainty. It discusses recent advances in qualitative, semi-quantitative, and quantitative assessment approaches, while also highlighting major challenges such as the parameterization of horizontal gene transfer, the integration of antimicrobial selection pressure, and the reliability of genomic-context evidence for resistance. Representative applications include WGS-based quantitative microbial risk assessment of foodborne pathogens, such as the Listeria monocytogenes case in which strain-level genomic heterogeneity was used to refine hazard characterization and exposure assumptions. Looking ahead, the integration of artificial intelligence is expected to further improve high-quality genome reconstruction, resistance phenotype inference, and transmission risk prediction, thereby facilitating more precise risk management.