Measured acceleration · Biology
BacterAI accelerates microbial metabolism mapping through active learning
Active-learning experiments inferred amino-acid requirements for oral streptococci with far fewer tests than exhaustive screening.
Summary
The Nature Microbiology paper introduced BacterAI as an active-learning system for microbial metabolism. It selected experiments to infer growth requirements for oral streptococci and reported major reductions in the fraction of the search space that needed to be experimentally tested.
AI role
An active-learning agent selected informative microbial growth experiments and updated hypotheses about metabolic requirements.
Narrative role
This backfills a 2023 biology example where AI changed the experimental search loop itself, adding historical depth beyond recent autonomous-lab materials examples.
Caveat
The demonstrated task is a specific microbial metabolism-mapping problem, so the acceleration claim is bounded to comparable combinatorial biology screens.