Machine Learning Approaches Using High-Throughput Profiling Data for Antibiotic Discovery

Clinical Snapshot

Key Findings

• P Value: Not reported at review level

• Effect Size: Not applicable — no quantitative pooled effect size reported; individual primary study performance metrics referenced descriptively

• Primary Outcome: Narrative synthesis demonstrating that ML applied to high-throughput profiling data (morphological phenotyping, Tn-seq, transcriptomics, proteomics, metabolomics) can systematically identify antibacterial hits and predict mechanisms of action for novel compounds in early-stage discovery

• Nnt Or Sensitivity: Not applicable at this stage; review addresses preclinical discovery metrics (model accuracy, MoA classification performance) rather than clinical NNT or diagnostic sensitivity/specificity

• Confidence Interval: Not reported — narrative review design precludes confidence interval estimation

Clinical Application

Implementation of ML-integrated high-throughput profiling requires substantial computational infrastructure, large curated datasets, interdisciplinary expertise (microbiology, bioinformatics, cheminformatics), and significant capital investment. Feasibility is currently limited to well-resourced academic research institutions and large pharmaceutical companies. The review does not address barriers to adoption, data sharing frameworks, or open-source tool availability. Australia faces significant AMR burden, with the ACSQHC and the Australian Government's National Antimicrobial Resistance Strategy 2020–2025 identifying novel antibiotic development as a priority. The Therapeutic Goods Administration (TGA) has adopted expedited pathways for novel antibiotics addressing unmet clinical need. Australian research institutions (e.g., Monash University, WEHI, University of Queensland) are active in computational drug discovery. PBS listing of novel antibiotics emerging from ML-accelerated pipelines would require standard Phase I–III clinical trial evidence regardless of discovery methodology. RACGP and ACSQHC antimicrobial stewardship guidelines would govern eventual clinical use. This review is most directly relevant to Australian researchers in the AMR discovery space rather than to frontline clinicians. Primarily relevant to pharmaceutical researchers, medicinal chemists, and microbiologists engaged in early-stage antibiotic discovery. Indirect relevance to infectious disease clinicians, clinical microbiologists, and antimicrobial stewardship teams who will ultimately utilise any antibiotics emerging from these pipelines. Patients with infections caused by multidrug-resistant organisms (MDROs) represent the ultimate beneficiary population.

Abstract

Antibiotic-resistant bacterial infections continue to increase globally, creating an urgent need for new antibacterials with novel mechanisms of action. Early stages of antibiotic discovery are often limited by the difficulty of identifying compounds that act through previously unrecognized pathways and by challenges in determining their mechanisms. Machine learning (ML) integrated with high-throughput profiling now provides systematic approaches to overcome these barriers. Beyond initial hit discovery, multilayer profiling using morphological phenotyping, transposon sequencing, transcriptomics, proteomics, and metabolomics captures cellular responses that reflect the mechanisms of action. Because profiling data sets are typically high-dimensional and contain defined features and variables, ML can extract complex patterns associated with pathway-level responses and predict mechanisms for unknown compounds. In this review, we summarize current progress in high-throughput profiling and describe how ML applied to each data set can accelerate the identification of antibacterials with new mechanisms. These approaches accelerate the transition from large-scale compound screening to mechanistic validation and enable effective prioritization of lead compounds in early-stage antibiotic discovery.

References

1. 1.Kim, I., Jang, J., Yun, T., Shin, Y., & Lee, W. (2026). Machine learning approaches using high-throughput profiling data for antibiotic discovery. ACS Infectious Diseases. Advance online publication. https://doi.org/10.1021/acsinfecdis.6c00004