An AI tool for prioritizing candidate biomarkers from wearable sensor data

Summary

Researchers at Google Research introduce the Biomarker Discovery Framework, a multi-agent system that prioritizes candidate biomarkers from wearable sensor data through iterative hypothesis generation, statistical analysis, adversarial validation, and literature-grounded reasoning, under human supervision. The system combines deterministic computation for numerical analysis with generative reasoning for hypothesis formation and interpretation, using an Orchestrator agent to guide specialized agents through a six-phase process. This approach addresses the limitations of existing language model-based systems that often overlook statistical validity on physiological time-series data. In three cohorts (N = 9,279 participant-observations), the framework recovered known clinical signals, identified convergent biomarkers across independent datasets, and improved downstream prediction when combined with demographic features. The framework maintains strict statistical rigor and preserves traceability through shared memory and structured fact sheets.