Decoding the Molecular Language of Proteins with Evolla
Interactive protein-language model answering natural-language queries over sequence and structure.

Xibin Zhou (周禧彬)
PhD candidate at Westlake University · AI for Science · Structure-aware PLMs & multimodal protein intelligence
View ResearchI am a Ph.D. candidate in Computer Science at Westlake University (joint program with Zhejiang University), based in Hangzhou. My research focuses on AI for Science, especially protein language models (PLMs) and their applications in protein engineering, search, and design.
I develop large-scale models and open platforms that bridge protein sequence, structure, and function—see Research for an overview, or Publications / Google Scholar for the full list.
* equal contribution
Interactive protein-language model answering natural-language queries over sequence and structure.
16B-parameter framework for instruction-driven structure generation and sequence design.
PLM-guided optimization enables programmable T→G/C base editing with few-shot validation.
ESM-1b semantic mining discovers multicopper oxidases with superior catalytic and environmental-remediation properties from UniProt.
Trimodal contrastive model unifying sequence, structure, and function text for billion-scale protein search.
No-code Colab platform for training, sharing, and collaborating on protein ML models.
General-purpose PLM with structure-aware vocabulary trained on ~40M sequence–structure pairs.
Evolla under review at Nature (bioRxiv preprint)
Pinal under review at Nature (bioRxiv preprint)
Best Poster at the Nature Conference on AI Augmented Biology (Nanjing University) — [poster]
ProTrek published in Nature Biotechnology
SaprotHub published in Nature Biotechnology
ESM-Ezy published in Nature Communications
SaProt accepted at ICLR 2024
Previous president of Water Sports Club in Westlake University