Reproduce “Constrained Diffusion for Protein Design with Hard Structural Constraints” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Jacob K. Christopher, Austin Seamann, Jingyi Cui, Sagar Khare, Ferdinando Fioretto

The authors introduce a diffusion method for generating protein backbones that satisfy local geometry and long-range constraints. At each step, it corrects a predicted clean structure, then adds noise; consensus ADMM coordinates local and global corrections. The method produces 21.0% usable PDZ designs versus zero for three baselines, and 97.8% usable encapsulation samples, though box-width violations were unreported.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.