Science through Computation Initiative
We are a group of 15 computational natural scientists working across the UK and meeting in London - united by a mission to accelerate scientific discovery through computational methods and tools. Founded in 2024, our collective brings together experts from chemistry, biology, computer science, and physics.
We focus on computational motifs – shared patterns and challenges that appear across scientific disciplines. By identifying and solving these core bottlenecks, we can create impact at scale, accelerating discovery across multiple fields simultaneously. This interdisciplinary approach allows us to spot opportunities that might be missed when working within traditional boundaries.
Our Mental Framework
Our approach systematically breaks down the scientific pipeline into three core components: exploration, collaboration, and experimentation, examining how computational advances can transform each stage. We analyze fundamental bottlenecks and identify computational solutions, developing both practical tools for immediate acceleration and novel algorithms that enable entirely new ways of conducting science.
This dual focus - on both enhancing existing processes and pioneering new computational methodologies - allows us to work toward both immediate improvements and transformative breakthroughs in how computational science is conducted across disciplines.
Events & Workshops
London Workshop
Following on the success of the San Francisco workshop, we're bringing a hackathon to London focused on the very same topic of accelerating scientific discovery with computation. Registration is open, sign up below and join us.
Building on the momentum, we'll tackle the same themes—addressing computational bottlenecks and advancing science through new innovative tools.
The London event will feature presentations from leading scientists and engineers, combined with hands-on coding sessions to create a space for real-time collaboration and experimentation. We'll share insights on computational challenges across scientific disciplines and demonstrate open-source tools that speed up experimentation.
We're looking to connect with UK academics, innovators, and organizations passionate about computational science. Interested in partnering or sponsoring? Please reach out—we'd love to explore collaboration opportunities.
San Francisco Workshop
This March, we brought our team to the Bay Area for a 2-day workshop focused on computational approaches to accelerating scientific discovery. Hosted by our good friends at pebblebed.
It was wonderful to bring the academic and tooling communities together to accelerate science. A quick summary of some of the day's projects:
- GreatProtocols (Winner!): AI-driven documentation tool ensuring complete and precise recording of experimental procedures.
- LikeMinds: Discover collaborators or competitors based on similar scientific interests from BlueSky likes.
- Deep-Critic: An LLM ensemble for reviewing academic papers.
- BYOL Hierarchy: Creation of hierarchical BYOL embeddings for long-range reasoning.
- MINERVA: Automated generation of causal graphs offering alternative hypotheses for observed outcomes.
- Basemotif: A team chat app which connects to scientific datasets, experiment orchestration, and literature.
The winners are joining us for a UK edition of this hackathon!
Who We Are
Interested in our work? We'd love to hear from you.
Kristina Kordova
University of Cambridge
Biology
Michał Wójcik
University of Oxford
Neuroscience
Puria Radmard
University of Cambridge
Neuroscience
Eva Sevenster
University of Bristol
Neuroscience
Austin Mroz
Imperial College London
Chemistry
Dan Akarca
Imperial College London
Neuroscience
Jakub Lála
Imperial College London
Biology
Martin Buttenschoen
University of Oxford
Biology
Filip Szczypiński
Durham University
Chemistry
Jonathon Langford
Imperial College London
Physics