Proteo-R1: Reasoning Foundation Models for De Novo Protein Design

TL;DR

Proteo-R1 introduces a dual-expert framework combining reasoning and generative models for more interpretable and controllable de novo protein design, explicitly reasoning about key residues before geometric synthesis.

Contribution

It presents a novel dual-expert architecture that separates understanding from generation, enabling explicit residue-level reasoning in protein design.

Findings

Achieves stable and interpretable protein design by decoupling reasoning from geometric generation.

Utilizes a multimodal large language model for residue analysis and a diffusion model for constrained geometry synthesis.

Code, data, and demos are publicly available at the provided URL.

Abstract

Deep learning in \emph{de novo} protein design has achieved atomic-level fidelity. However, existing models remain largely non-deliberative: they directly synthesize molecular geometries without explicitly reasoning about which residues or interactions are functionally essential. As a result, design decisions are entangled with continuous sampling dynamics, limiting interpretability, controllability, and systematic reuse of biochemical knowledge. We introduce \textbf{Proteo-R1}, a reasoning-guided protein design framework that explicitly decouples \emph{molecular understanding} from \emph{geometric generation}. Proteo-R1 adopts a dual-expert architecture in which a multimodal large language model (MLLM) serves as an \emph{understanding expert}, analyzing protein sequences, structures, and textual context to identify key functional residues that govern binding and specificity. These residue-level decisions are then passed as hard constraints to a separate diffusion-based \emph{generation expert}, which performs conditional co-design while respecting the fixed interaction anchors. This factorization mirrors how human experts approach molecular engineering: first, reasoning about critical interactions, then optimizing geometry subject to those constraints. By operationalizing reasoning as explicit residue-level commitments rather than latent textual guidance, Proteo-R1 achieves stable, interpretable, and modular integration of LLM reasoning with state-of-the-art geometric generative models. Code, data, and demos are available at https://smiles724.github.io/r1/.