BioDyad synchronizes biomedical discovery and machine learning engineering
BioDyad: Synchronize Biomedical Discovery and Machine Learning Engineering
Artificial Intelligence
Summary
Biomedical research often needs computer programs that can learn from data and also use the latest scientific knowledge. The authors created BioDyad, a system that helps combine new biomedical discoveries with the building and testing of machine learning programs. It carefully manages how evidence guides program creation and how the programs get trained and checked step-by-step. They tested BioDyad on many biomedical tasks and found it worked better than other similar methods, showing that linking discovery and engineering can improve biomedical AI.
What this means in practice
- •For biomedical data scientists: Automate building and validating machine learning models that incorporate scientific knowledge across multiple biomedical tasks within fixed time limits.
- •For pharmaceutical analytics teams: Use BioDyad to better integrate evolving biomedical evidence into drug discovery modeling pipelines, improving predictive accuracy and efficiency.
Authors
Xingbo Du, Fadli Aulawi Al Ghiffari, Leonard Song, Loka Li, Duzhen Zhang, Zixiao Wang, Xiuying Chen, Le Song
Abstract
Agentic biomedical machine learning (ML) draws on complementary advances in biomedical evidence acquisition and executable program search. Existing systems connect aspects of these capabilities, but coordinating them throughout program search remains challenging. New evidence must guide candidate construction, execution outcomes must inform subsequent discovery and reuse, and validation demands must fit the search budget. We introduce BioDyad, which couples biomedical discovery and ML engineering through two hierarchies within Monte Carlo graph search. Its scientific hierarchy combines prior biomedical guidance with iterative discovery, then links biomedical plans to execution outcomes in memory for reuse across candidates. Its engineering hierarchy moves candidate programs from smoke execution, through train/validation evaluation, to full-data retraining. We evaluate BioDyad on the 76-task BioXArena benchmark under a two-hour per-task budget with three matched LLM backends. It achieves the highest penalized all-task score and task success rate among four agent methods and a one-shot baseline under each backend. These results support coordinating biomedical discovery and ML engineering to integrate external knowledge into executable programs across heterogeneous biomedical tasks.