Subjects

- Machine learning

- Medical research

Abstract

Scientific discovery is driven by scientists generating novel hypotheses for complex problems that undergo rigorous experimental validation. To augment this process, we introduce Co-Scientist, a multi-agent AI system built on Gemini for structured scientific thinking and hypothesis generation. Co-Scientist aims to help scientists discover new original knowledge. Conditioned on their research objectives and prior scientific evidence, it formulates demonstrably novel research hypotheses for experimental verification. The system’s design involves agents continuously generating, critiquing and refining hypotheses accelerated by scaling test-time compute. Key contributions include: (1) a multi-agent architecture with an asynchronous task execution framework for flexible compute scaling; (2) a tournament evolution process for self-improving hypotheses generation. Automated evaluations show continued benefits of test-time compute scaling, improving hypothesis quality over time. While general purpose, we focus the validation in three biomedical applications: drug repurposing, novel target discovery 1, and explaining mechanisms of anti-microbial resistance 2. Specifically, Co-Scientist helped identify new drug repurposing candidates and synergistic combination therapies for acute myeloid leukemia, which were validated through in vitro experiments. These real-world validations demonstrate the potential of Co-Scientist to accelerate scientific discovery and usher in an era of AI empowered scientists.

You have full access to this article via your institution.

Download PDF

Author information

Author notes- These authors contributed equally: Juraj Gottweis, Wei-Hung Weng, Alexander Daryin, Tao Tu, Petar Sirkovic, Artiom Myaskovsky, Grzegorz Glowaty, Felix Weissenberger, Alessio Orlandi

Authors and Affiliations

- Google Cloud AI Research, Zurich, Switzerland

Juraj Gottweis, Alexander Daryin, Petar Sirkovic, Artiom Myaskovsky, Grzegorz Glowaty, Felix Weissenberger, Alessio Orlandi, Dina Zverinski, Ivor Rendulic, Florian Hasler, Luka Rimanic, Marina Boia, Ivan Budiselic, Burak Gokturk & Amin Vahdat

- Google DeepMind, Mountain View, California, USA

Wei-Hung Weng, Tao Tu, Keran Rong, Ryutaro Tanno, Khaled Saab, Nenad Tomašev, Elahe Vedadi, Jan Freyberg, Jeremy Ratcliff, Ottavia Bertolli, Demis Hassabis, Yunhan Xu, Pushmeet Kohli, Alan Karthikesalingam & Vivek Natarajan

- Google Research, Mountain View, California, USA

Dan Popovici, Anil Palepu, Fan Zhang, Andrew Carroll, Kavita Kulkarni, Ben Feinstein, Mathias Bellaiche, Tom Sheffer, Katherine Chou, Avinatan Hassidim, Yossi Matias, James Manyika & Annalisa Pawlosky

- Stanford University School of Medicine, Palo Alto, California, USA

Jacob Blum, Yuan Guan & Gary Peltz

- Houston Methodist, Houston, Texas, USA

Vikram Dhillon

- Sequome, South San Francisco, California, USA

Eeshit Dhaval Vaishnav & Byron Lee

- Fleming Initiative and Imperial College London, London, UK

Tiago R. D. Costa & José R. Penadés

Authors- Juraj GottweisView author publications