- Run and benchmark open-source protein design methods on real Adaptyv targets, validated against experimental data from our wet lab.
- Design binders for internal R&D campaigns, and track experimental performance across success rate, hit rate, kinetics, and developability.
- Build computational tooling around the design pipeline: structure prediction, filtering, and ranking, to triage large design pools down to candidates for synthesis and characterization.
- Support the technical side of our open competitions and hackathons: drafting track briefs, supporting participants, judging submissions, and writing up results.
- Contribute open data, blog posts, designer spotlights, and method comparisons to Proteinbase.
PythonMachine LearningPyTorch