A single-stage AI can now design the entire atomic structure of a protein at once, including synthetic parts that do not exist in nature.
Traditional protein design is a slow, two-part process that first picks a shape and then tries to find a matching sequence. This new diffusion model handles both the sequence and the physical coordinates of every atom simultaneously. This unified approach allows researchers to incorporate non-canonical amino acids that are entirely synthetic. Designing proteins with these alien building blocks could lead to ultra-stable enzymes for industry or completely new types of medicines. It moves us past the limited palette of 20 natural amino acids that has governed life for billions of years.
A-CODE: Fully Atomic Protein Co-Design with Unified Multimodal Diffusion
arXiv · 2605.03360
We present A-CODE, a fully atomic unified one-stage protein co-design model that simultaneously refines discrete atom types and continuous atom coordinates. Unlike predominant two-stage methods that cascade structure design with amino acid-level sequence design, our approach is fully atomic within a unified multimodal diffusion framework, in which residue identities are inferred solely from atom-level predictions. Built upon the powerful all-atom architecture, A-CODE achieves superior designabil