Overview
Antimicrobial resistance (AMR) is one of the most urgent challenges in global health, and much of AI’s potential to help remains untapped because the AI-ready data and evaluations needed to build predictive models don’t yet exist. Align partnered with Google DeepMind to change that, convening the global AMR community to define and prioritize the datasets that could power the next generation of predictive models for drug resistance.
Through an open call for proposals and community ideation sessions, researchers, clinicians, and technologists across the AMR space worked together to surface the field’s highest-priority “learnable” problems. That process produced a landscape report of community priorities and five technically defined, funding-ready dataset and model proposals, each built with an expert team ready to execute. This work builds on the calls to action in the AI for AMR report from Google DeepMind and the Fleming Initiative.
Five Priority Proposals
Community priorities turned into five funding-ready dataset and model proposals, each ready for collaborators, data contributors, and funders.
- Predicting Quantitative MIC from Whole Genome Sequencing
- Predicting Permeability & Efflux of Antibiotic Compounds
- Predicting Treatment-Emergent AMR at the Patient Level in LMICs
- Data to Dose: AI-Guided Validation of Pediatric Antibiotic Regimens
- AI-Enabled Mechanistic Toxicity Prediction for Antibiotic Discovery