NEW Phage Foundry Preprint: Machine Learning (ML) Framework to Predict Strain-level Phage-host Interactions

The Phage Foundry team is excited to share a new preprint led by Phage Foundry postdoctoral researcher Avery Noonan (LBNL, Arkin Lab), with support from postdoctoral researcher Lucas Morinière (LBNL, Mutalik & Arkin Lab), a colleague working on the NSF EDGE project.

The team developed a machine learning (ML) framework to predict strain-level phage-host interactions across diverse bacterial genera from genome sequences alone. This work serves as an outstanding example of how teams can conduct rigorous AI/ML work in phage biology, genotype-phenotype and therapy field.

Noonan optimized 13.2 million training runs across five datasets (128,357 interactions, 1,058 strains, 560 phages), demonstrating great performance while eliminating phylogenetic constraints. Morinière led all experimental validations. The model-guided cocktail design achieved 80.7-97.0% strain coverage w 5 phages.

This platform enables rational phage therapy design and precision microbiome engineering with applications across clinical, agricultural, and industrial contexts.

This work was supported through DOE BRaVE initiative for Berkeley Lab’s Phage Foundry work and the NSF EDGE program.