Trion 1 unifies protein language sequence and structure modeling
Trion-1: Unifying Language, Protein Sequence, and Structure with Diffusion Language Models
Artificial Intelligence
Summary
Proteins are made of sequences that fold into structures determining their functions, but understanding and designing them is complicated. The authors created Trion-1, a new model that combines protein sequences, their 3D structures, and natural language all in one system. Trion-1 learns the patterns and relationships in these different data types simultaneously, allowing it to predict how proteins fold, design new sequences, and interpret related language. The model performs well compared to existing methods, showing promise for more integrated protein research tools.
What this means in practice
- •For biotech engineering teams: Generate and design protein sequences and structures simultaneously for novel therapeutic proteins or enzymes using an integrated multimodal model.$Commercial implications: Offers a new tool to design and fold proteins more efficiently, enabling commercial development of custom proteins for industry and medicine.
- •For bioinformatics platform developers: Build protein analysis tools that interpret sequence and structure along with related scientific language for better protein function prediction and annotation.
Authors
Tianyu Lu, Kaicheng Yang, Jiangtao Feng, Keyue Qiu, Yawen Ouyang, Yixin Wu, Xinbo Zhang, Lihao Wang, Wei-Ying Ma, Hao Zhou
Abstract
We envision a general-purpose protein foundation model as a unified, interactive model of sequence, structure, and function, capable of understanding, generation, and cross-modal translation while retaining the instruction-following and reasoning capabilities of modern LLMs. Toward this vision, we introduce Trion-1, the first native, unified, multimodal diffusion language model (dLLM) for proteins. Trion-1 is built on two key design choices: (1) a unified discrete token space that jointly represents sequence, structure, and natural language; (2) a dLLM backbone that replaces strict left-to-right factorization with bidirectional or semi-bidirectional token interactions, enabling global residue communication and long-range contact modeling that are naturally suited to structure modeling. This formulation unifies six core capabilities: folding, inverse folding, sequence understanding, structure understanding, sequence design, and structure design, while retaining the native language capabilities of general-purpose LLMs. On established benchmarks, Trion-1 achieves broadly competitive performance. In folding, Trion-1 achieves a TM-score above 0.8 with a single sample under strict homology control, matching or exceeding the existing tri-modal model using 200 samples. For sequence understanding, Trion-1 performs on par with top-tier LLMs such as Claude Opus 4.7 (62.37%). These results support well-aligned sequence-structure-language representations and strong cross-modal generative capabilities.