MADD: Multi-Agent Drug Discovery Orchestra
Gleb V. Solovev, Alina B. Zhidkovskaya, Anastasia Orlova, Nina Gubina, Anastasia Vepreva, Rodion Golovinskii, Ilya Tonkii, Ivan Dubrovsky, Ivan Gurev, Dmitry Gilemkhanov, Denis Chistiakov, Timur A. Aliev, Ivan Poddiakov, Galina Zubkova, Ekaterina V. Skorb, Vladimir Vinogradov, Alexander Boukhanovsky, Nikolay Nikitin, Andrei Dmitrenko, Anna Kalyuzhnaya, Andrey Savchenko
2025-11-13
Summary
This paper introduces a new AI system called MADD that helps researchers find potential drug candidates more easily and effectively.
What's the problem?
Traditionally, finding the first promising molecules, called 'hits', for a new drug is a slow and expensive process that requires a lot of lab work. While recent AI tools, especially those using large language models, can speed things up, they've become complicated and hard for scientists who primarily work in the lab to use directly.
What's the solution?
The researchers created MADD, which is like a team of AI 'agents' working together. You can simply describe what kind of molecule you're looking for in plain language, and MADD automatically builds and runs a customized search process to find potential hits. It combines the ability of language models to understand your request with the precision of specialized tools for designing and testing molecules. They tested MADD on several real-world drug discovery problems and showed it performs better than other AI-based methods.
Why it matters?
MADD makes advanced AI drug discovery tools more accessible to researchers without extensive AI expertise, potentially accelerating the development of new medicines. They also created a new standard dataset to help others build and improve similar AI systems, pushing the field forward towards a future where AI plays a bigger role in designing drugs.
Abstract
Hit identification is a central challenge in early drug discovery, traditionally requiring substantial experimental resources. Recent advances in artificial intelligence, particularly large language models (LLMs), have enabled virtual screening methods that reduce costs and improve efficiency. However, the growing complexity of these tools has limited their accessibility to wet-lab researchers. Multi-agent systems offer a promising solution by combining the interpretability of LLMs with the precision of specialized models and tools. In this work, we present MADD, a multi-agent system that builds and executes customized hit identification pipelines from natural language queries. MADD employs four coordinated agents to handle key subtasks in de novo compound generation and screening. We evaluate MADD across seven drug discovery cases and demonstrate its superior performance compared to existing LLM-based solutions. Using MADD, we pioneer the application of AI-first drug design to five biological targets and release the identified hit molecules. Finally, we introduce a new benchmark of query-molecule pairs and docking scores for over three million compounds to contribute to the agentic future of drug design.