Demystify Protein Generation with Hierarchical Conditional Diffusion Models

TL;DR

The paper tackles de novo protein design under functional constraints by introducing a multi-level conditional diffusion framework that jointly generates amino acid sequences, backbone geometries, and all-atom coordinates in a SE(3)-invariant, end-to-end manner. A conditional flow links the three hierarchical levels, enabling faithful conditioning on target functions, while the Protein-MMD metric provides a principled way to assess distributional and functional consistency using ESM2 embeddings. Across EC and GO benchmarks, the approach yields strong conditional consistency and competitive structural diversity, with ablations confirming the necessity of all three levels. This work offers a scalable design pathway for function-guided protein generation and a robust evaluation tool for conditional protein design.

Abstract

Generating novel and functional protein sequences is critical to a wide range of applications in biology. Recent advancements in conditional diffusion models have shown impressive empirical performance in protein generation tasks. However, reliable generations of protein remain an open research question in de novo protein design, especially when it comes to conditional diffusion models. Considering the biological function of a protein is determined by multi-level structures, we propose a novel multi-level conditional diffusion model that integrates both sequence-based and structure-based information for efficient end-to-end protein design guided by specified functions. By generating representations at different levels simultaneously, our framework can effectively model the inherent hierarchical relations between different levels, resulting in an informative and discriminative representation of the generated protein. We also propose a Protein-MMD, a new reliable evaluation metric, to evaluate the quality of generated protein with conditional diffusion models. Our new metric is able to capture both distributional and functional similarities between real and generated protein sequences while ensuring conditional consistency. We experiment with the benchmark datasets, and the results on conditional protein generation tasks demonstrate the efficacy of the proposed generation framework and evaluation metric.

Figures (8)

• Figure 1: Protein visualization.
• Figure 2: The architecture of the multi-level diffusion model.
• Figure 3: Consistency in the latent space.
• Figure 4: Consistency in the latent space
• Figure 5: Consistency.