Genetic algorithms help multiple ai agents find better scientific hypotheses
HypoEvolve: Genetic Algorithms Enable Multi-Agent LLMs to Discover Scientific Hypotheses
Computation and LanguageComputational Engineering, Finance, and ScienceMultiagent SystemsNeural and Evolutionary Computing
Summary
Finding good scientific explanations often needs people to work together, check ideas, and use evidence. The authors created HypoEvolve, a system that uses many large language AI models working like a team guided by genetic algorithms to improve scientific hypotheses across generations. They tested it by seeing how well it explains drug effects for different cancers, using biological data to evaluate the ideas. HypoEvolve did better than other systems, showing that multiple AI agents collaborating can discover stronger scientific ideas than working alone.
What this means in practice
- •For biotech research teams: Generate and refine hypotheses about drug mechanisms across multiple diseases by coordinating AI agents to integrate evidence and test new explanations.
- •For pharmaceutical data teams: Assess candidate drug repurposing hypotheses with a collaborative AI system that improves explanation quality based on genetic and clinical data.
Authors
Jieyuan Liu, Mengzhou Hu, Jefferson Chen, JungHo Kong, Pratibha Jagannatha, Yiming Gao, Dexter Pratt, Hsin-Yuan Lee, Zhiting Hu, Trey Ideker, Wei Wang, Eric P. Xing, Zhen Wang
Abstract
Scientific agents contribute to hypothesis discovery by synthesizing evidence, assessing proposals, and developing new explanations. Recent systems combine scientific agents with evolutionary search through critique, comparison, and revision. However, how different forms of agent collaboration affect hypothesis quality remains an open question. Answering this question requires separating the effects of agents' scientific capabilities from those of their collaboration. A framework must therefore preserve agents' scientific roles and support rules for combining, revising, and retaining hypotheses. Building on this view, we introduce HypoEvolve, which makes collaboration explicit through successive updates to a hypothesis population. Specifically, we propose a generational genetic algorithm to coordinate specialized large language model (LLM) agents that integrate mechanistic arguments, reconsider assumptions, and assess evidence and testability. Each generation specifies how scientific judgments and new proposals reshape the population, making collaboration effects on hypothesis quality directly testable. Moreover, we design our evaluation around scientifically meaningful hypotheses that explain how a proposed intervention could work. Drug repurposing links these explanations to target-level biological claims assessed against external evidence. Specifically, we adapt DepMap and Open Targets into complementary external measures grounded in experimental, genetic, and clinical evidence. Across 34 cancer types, HypoEvolve achieves the highest scores against six baselines on both measures. DepMap selectivity reaches 0.171, versus 0.115 for the strongest baseline. Gains over single-pass generation also generalize to held-out cancer types. HypoEvolve advances a vision of autonomous science in which AI research teams achieve a capacity for discovery beyond that of individual models.