Co-op, Agentic and Generative Protein Design

Your Impact at LILA

Our team works on AI methods for therapeutic protein design, with a tight loop between computation and the wet lab. A co-op could take on either of two tracks, or a mix of both. (1) Generative models and protein design pipelines: build, benchmark, and improve methods for designing proteins (such as antibodies and minibinders), with candidates tested in our lab so results are grounded in experimental data. (2) AI agents for protein design campaigns: build and evaluate agents that help plan campaign strategy (target and epitope selection, choosing methods and settings), execute campaigns (launching design and filtering runs, diagnosing failures, adapting the next round), and reason over biological priors from the literature alongside the outputs of domain-specific models. Either way, the co-op would own a well-scoped project, ship code and analyses that feed real design decisions and scientific outcomes.

What You'll Be Building

• Develop, evaluate, and apply agentic systems for protein design and/or deep generative models for protein design.

What You'll Need to Succeed

• Strong python and and computational science background.
• Experience with coding agents.
• Experience with machine learning for molecular design

Bonus Points For

• Experience with customizing agents, building harnesses, or using agents for scientific applications
• Solid technical understand and experience with generative models.