Background: Antimicrobial resistance (AMR) poses a critical threat to global health, driven by the rapid evolution of pathogenic microorganisms and the widespread misuse of antibiotics. Conventional culture-based antimicrobial susceptibility testing is time-consuming and often inadequate for detecting emerging resistance mechanisms. Advances in whole-genome sequencing, combined with bioinformatics and machine learning (ML), offer promising alternatives for rapid and accurate resistance prediction. Methods: This discussion-based review synthesizes peer-reviewed studies published between 2010 and 2024 that applied bioinformatics pipelines and ML algorithms to genome-based AMR prediction. Data sources included whole-genome sequencing datasets, curated resistance databases, and phenotypic susceptibility profiles. Results: ML models demonstrated high accuracy (up to 94%) in predicting resistance phenotypes across major bacterial pathogens. Deep learning approaches showed superior sensitivity in identifying complex and novel resistance determinants. Integration of genomic prediction into clinical workflows significantly reduced diagnostic turnaround time and supported improved antimicrobial stewardship. Conclusion: Bioinformatics-driven ML frameworks provide powerful tools for addressing AMR through rapid resistance prediction and enhanced surveillance. Despite challenges related to data bias and interpretability, continued methodological advancements and standardized data practices are likely to accelerate clinical adoption and contribute meaningfully to combating antimicrobial resistance. References: Antimicrobial resistance, Bioinformatics, Machine learning