Published in NeurIPS 2026 (Under review), 2026
KnowEvo turns experimental feedback into reusable, executable knowledge that evolves through repeated evaluation and revision for protein optimization.
Published in Information Processing & Management, 2026
S2Token treats chemically meaningful molecular substructures as tokens, improving both representativeness and generalization for molecular large language models.
Recommended citation: Runze Wang, Zijie Xing, Xingyue Liu, Mingqi Yang, Che He, Yanming Shen*, “From Graphs to Tokens: Substructure-Aware Molecular Representation for Large Language Models,” Information Processing and Management, vol. 63, no. 6, p. 104771, Sep. 2026
Download Paper | Download Bibtex
Published in ICLR 2026 Workshop on AI with Recursive Self-Improvement, 2026
This work reframes protein optimization as evolving an explicit expertise pool, using language-guided multi-agent feedback instead of parameter-heavy reinforcement learning.
Recommended citation: Xingyue Liu†, Zijie Xing†, Runze Wang, Luoming Hu, and Yanming Shen*, “Language-guided expertise evolution for protein optimization,” ICLR 2026 Workshop on AI with Recursive Self-Improvement, 2026
Download Paper | Download Slides/Poster | Download Bibtex
Published in Expert Systems with Applications, 2025
Uni-Motif integrates molecular motifs into GNNs through decoupled functional and structural encodings, improving expressiveness across different architectures.
Recommended citation: Runze Wang, Yuting Ma, Xingyue Liu, Zijie Xing, and Yanming Shen*, “Motif-driven molecular graph representation learning,” Expert Systems with Applications, vol. 269, p. 126484, 2025
Download Paper | Download Bibtex