Protein & Binder Design

Computational approaches to de novo protein and antibody design using diffusion models, structure prediction, and generative AI. I participate in open binder design competitions where designs are experimentally validated in the wet lab.

For my antibody design research (benchmarks, tools, analysis), see the Antibody Research hub.

Competitions

Open binder design challenges with experimental validation. Designs are synthesized and tested in the lab by the competition organizers.

Resources & Reading

A curated collection of the best writing on antibody engineering, protein design, and computational biology. Useful starting points whether you're new to the field or deep in the weeds.

Primers & Overviews

Designing Antibodies with AI — Asimov Press

Accessible overview of how AI is reshaping antibody discovery, from structure prediction to generative design.

How to Make an Antibody — Works in Progress

The full journey of antibody development from biology to manufacturing, written for a general technical audience.

AI Antibody Design: State of the Art (2025) — Brian Naughton, Boolean Biotech

Comprehensive technical review of every major AI antibody design platform and method as of 2025.

Owl Posting — Antibody Engineering Series

A Primer on Machine Learning in Antibody Engineering

Antibody structure basics (Fab, CDRs, variable regions), key datasets like OAS, and a walkthrough of major ML models in the field.

A Primer on the Next Generation of Antibodies

What's broken about traditional antibodies, and three alternatives: scFvs, nanobodies (VHH), and antibody mimetics.

Better Antibodies by Engineering Targets, Not Antibodies (Nabla Bio)

Deep-dive on Nabla Bio's approach: engineer the target presentation, not just the antibody, to get better binders.

Escalante Bio — Practical Protein Design

Minibinder Design Is Just Not That Hard

A frank take on the current state of computational binder design—the tools work, and the barrier to entry is lower than you think.

180 Lines of Code to Win the In Silico Portion of the Adaptyv Nipah Competition

How a minimal pipeline won the computational round of the same Nipah competition I participated in.

Winning the De Novo Portion of the Adaptyv Nipah Binder Competition

The experimental results: which designs actually bound, and what that tells us about the in silico–to–wet lab gap.

Teaching Generative Models to Hallucinate

On mosaic, Escalante's JAX-based PSSM optimization tool for protein design, and coaxing diffusion models into novel folds.

Petri: Abstractions for Lab Automation

Software abstractions for bridging the gap between computational design and automated wet-lab execution.

Your Experiment Has a Runtime

Thinking about biological experiments through the lens of computational complexity—wall-clock time matters.

My Antibody Research

Analysis, benchmarks, and tools for de novo antibody design.

Antibody Research Hub → rfab-harness → DADB Benchmark → FLAb Analysis →