Biology comes first.
Start with meaningful scientific questions. Let the biological problem guide the work.
Approach
Designed around scientific questions.
Our aim is to develop AI systems for connected research problems, rather than isolated tasks.
01
Connected questions
A biological question rarely stands alone. It leads to further questions, invites different perspectives, and changes as the evidence develops.
We are interested in how computational tools can contribute to that process and to deeper biological understanding. This is the direction of our work on agentic research systems: AI that can support the process of scientific investigation.
02
Working principles
Evaluate with care.
Examine what a system can do, where it falls short, and what the evidence supports.
Make uncertainty clear.
Distinguish an observation from an interpretation, and a promising direction from a demonstrated result.
03
Principles in practice
Our open Molecular Animation framework applies the same discipline to scientific illustration.
Its films are built from deposited structures. The claims a film depends on are measured in code before anything is animated, and every beat in a labelled film states on screen whether it shows an experimental structure or an illustrative step.
See Molecular Animation04
What we share, and when
Our research systems are in development and are not publicly available.
We intend to share results when they are ready for scrutiny, together with their scope and limitations. Until then, we would rather describe our direction accurately than suggest more than the evidence supports.
Contact
Start a conversation.
For research collaborations, molecular animation, partnerships and press, we would be glad to hear from you.