BioDisco: Multi-Agent Hypothesis Generation with Dual-Mode Evidence, Iterative Feedback and Temporal Evaluation

Ke, Yujing; George, Kevin; Pandya, Kathan; Großmann, Gerrit; Blumenthal, David B.; Sprang, Maximilian; Selby, David A.; Vollmer, Sebastian

doi:10.24963/ijcai.2026/752

BioDisco: Multi-Agent Hypothesis Generation with Dual-Mode Evidence, Iterative Feedback and Temporal Evaluation

Yujing Ke, Kevin George, Kathan Pandya, Gerrit Großmann, David B. Blumenthal, Maximilian Sprang, David A. Selby, Sebastian Vollmer

Download BibTeX Cite

Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence. AI and Health. Pages 6761-6769.

https://doi.org/10.24963/ijcai.2026/752

Keywords

Biomedical NLP: Biomedical NLP
Medical knowledge representation: Medical knowledge representation
LLM in medicine: LLM in medicine
AI4H: Drug discovery

Abstract

Identifying novel hypotheses is essential to scientific research, yet this process risks being overwhelmed by the sheer volume and complexity of available information. Existing automated methods often struggle to generate novel and evidence-grounded hypotheses, lack robust iterative refinement and rarely undergo rigorous temporal evaluation for future discovery potential. To address this, we propose BIODISCO, a multi-agent framework that draws upon language model-based reasoning and a dual-mode evidence system (biomedical knowledge graphs and automated literature retrieval) for grounded novelty, integrates an internal scoring and feedback loop for iterative refinement, and validates performance through pioneering temporal and human evaluations and a Bradley-Terry paired comparison model for statistical assessment. Evaluations suggest improved novelty and significance relative to ablated configurations and a generalist biomedical agent. Designed for flexibility and modularity, BIODISCO allows seamless integration of custom language models or knowledge graphs, and can be run with just a few lines of code.

Yujing Ke, Kevin George, Kathan Pandya, Gerrit Großmann, David B. Blumenthal, Maximilian Sprang, David A. Selby, and Sebastian Vollmer. 2026. BioDisco: Multi-Agent Hypothesis Generation with Dual-Mode Evidence, Iterative Feedback and Temporal Evaluation. In Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26. IJCAI, 6761–6769. https://doi.org/10.24963/ijcai.2026/752

@inproceedings{ijcai2026p752,
  title     = {BioDisco: Multi-Agent Hypothesis Generation with Dual-Mode Evidence, Iterative Feedback and Temporal Evaluation},
  author    = {Ke, Yujing and George, Kevin and Pandya, Kathan and Großmann, Gerrit and Blumenthal, David B. and Sprang, Maximilian and Selby, David A. and Vollmer, Sebastian},
  booktitle = {Proceedings of the Thirty-Fifth International Joint Conference on
               Artificial Intelligence, {IJCAI-26}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Diego Calvanese},
  pages     = {6761--6769},
  year      = {2026},
  month     = {8},
  note      = {AI and Health},
  doi       = {10.24963/ijcai.2026/752},
  url       = {https://doi.org/10.24963/ijcai.2026/752},
}

TY  - CPAPER
TI  - BioDisco: Multi-Agent Hypothesis Generation with Dual-Mode Evidence, Iterative Feedback and Temporal Evaluation
AU  - Ke, Yujing
AU  - George, Kevin
AU  - Pandya, Kathan
AU  - Großmann, Gerrit
AU  - Blumenthal, David B.
AU  - Sprang, Maximilian
AU  - Selby, David A.
AU  - Vollmer, Sebastian
T2  - Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence, IJCAI-26
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2026
SP  - 6761
EP  - 6769
DO  - 10.24963/ijcai.2026/752
UR  - https://doi.org/10.24963/ijcai.2026/752
ER  -