# All-atom protein sequence design using discrete diffusion models

## TL;DR

This paper introduces a new method for designing protein sequences using all-atom representations, allowing for more diverse and novel protein designs.

## Contribution

The work introduces a discrete diffusion model using SELFIES for all-atom protein sequence design, enabling inclusion of non-canonical amino acids.

## Key findings

- All-atom models generate proteins with higher novelty and diversity compared to amino acid-based models.
- The absorbing noise schedule in diffusion models performs better than the uniform schedule.
- Generated proteins show structural foldability comparable to conventional models.

## Abstract

Advancing protein design is crucial for breakthroughs in medicine and biotechnology. Traditional approaches for protein sequence representation often rely solely on the 20 canonical amino acids, limiting the representation of non-canonical amino acids and residues that undergo post-translational modifications. This work explores discrete diffusion models for generating novel protein sequences using the all-atom chemical representation SELFIES. By encoding the atomic composition of each amino acid in the protein, this approach expands the design possibilities beyond standard sequence representations. Using a modified ByteNet architecture within the discrete diffusion D3PM framework, we evaluate the impact of this all-atom representation on protein quality, diversity, and novelty, compared to conventional amino acid-based models. To this end, we develop a comprehensive assessment pipeline to determine whether generated SELFIES sequences translate into valid proteins containing both canonical and non-canonical amino acids. Additionally, we examine the influence of two noise schedules within the diffusion process—uniform (random replacement of tokens) and absorbing (progressive masking)—on generation performance. While models trained on the all-atom representation struggle to consistently generate fully valid proteins, the successfully generated proteins show improved novelty and diversity compared to their amino acid-based model counterparts. Furthermore, the all-atom representation achieves structural foldability results comparable to those of amino acid-based models. Lastly, our results highlight the absorbing noise schedule as the most effective for both representations. Data and code are available at https://github.com/Intelligent-molecular-systems/All-Atom-Protein-Sequence-Generation.

The online version contains supplementary material available at 10.1186/s13321-025-01121-1.

This work introduces a discrete diffusion-based framework for protein sequence generation using an all-atom representation, laying the groundwork for extending to non-canonical amino acids and post-translational modifications. Additionally, it provides a comprehensive evaluation pipeline to assess the validity of generated proteins, demonstrating how noise schedules within the diffusion process impact sequence novelty, diversity, and structural foldability.

The online version contains supplementary material available at 10.1186/s13321-025-01121-1.