Skip to content
DraconiticResearch
Search Sources API
197,788 studies31 sourcesUpdated 42 min ago
Peer-reviewedCC BY 4.0Bioinformatics

Katalyst: knowledge-guided semantic alignment via contrastive learning for enzyme turnover prediction

  • Ruilin Li1,2
  • Senyu Tang1
  • Jiaqi Deng3
  • Taixing Qiu4
  • Fei Guo5
  • Jijun Tang2
  • Xiaoyi Liu6
6 affiliations
  1. 1College of Computer Science and Software Engineering, Shenzhen University , Shenzhen, 518060,
  2. 2School of Computer Science and Control Engineering, Shenzhen University of Advanced Technology , Shenzhen, 518107,
  3. 3Department of Computer Science and Software Engineering, Southern University of Science and Technology , Shenzhen, 518055,
  4. 4School of Computer Science, Nanjing University , Nanjing, 210023,
  5. 5School of Computer Science and Engineering, Central South University , Changsha, 410083,
  6. 6School of Chinese Materia Medica, Beijing University of Chinese Medicine , Beijing, 102488,
Open at sourcePDFDOICite

Abstract

Motivation: The turnover number (kcat) is crucial for understanding enzyme kinetics, yet its experimental determination remains costly and low-throughput. Existing deep learning approaches often rely on single-substrate assumptions and overlook full reaction contexts or language-informed semantic representations, limiting their accuracy and generalization across diverse reactions.

Results: To address these challenges, we propose Katalyst, a knowledge-guided semantic alignment framework for kcat prediction based on contrastive learning. Katalyst integrates enzyme sequences, protein textual descriptions, substrates and products into a shared latent space using multi-channel encoders and contrastive alignment. An interaction-aware fusion module is designed to capture complementary cross-modal information. Experimental results demonstrate that Katalyst achieved the best overall performance, with the RMSE of 0.9880, R2 of 0.5829, and SRCC of 0.7603. Compared with the State-Of-The-Art (SOTA) baselines, Katalyst reduced RMSE by 3.01% and improved R2 and SRCC by 4.72% and 2.31%, respectively. Moreover, ablation studies and evaluations across sequence scales further confirmed the effectiveness of semantic alignment and multimodal feature fusion. Overall, Katalyst provides a robust and accurate framework for kcat prediction.

Availability and implementation: The code and data are available at https://github.com/liruilin-SUAT/Katalyst.

Related

  1. EvoGraph: synergizing evolutionary semantics and tail-aware graph contrastive learning for protein-protein interaction prediction.
    Yilou Zhang, Yuan Zhao, Jue Zhou +2
    Briefings in bioinformaticsPeer-reviewedCC BY-NCFull textMolecular biology & gene regulationEurope PMCPMC
  2. Multimodal alignment improves generalizability of genomic biomarker prediction in computational pathology
    Ekaterina Redekop, Eric Zimmermann, Ava P. Amini +5
    Cell Reports MethodsPeer-reviewedCC BY-NC-ND 4.0Full textMolecular biology & gene regulationPMC
  3. AOXP: An Integrated Framework for Antioxidant Peptide Prediction Leveraging LLM-Based Semantic Features and Optimized Physicochemical Ensembles
    Leilei Zhang, Zhongao Lu, Xiaofei Zhou +2
    ACS OmegaPeer-reviewedCC BY-NC-ND 4.0Full textMolecular biology & gene regulationCrossref
  4. Explainable/interpretable artificial intelligence and knowledge graph neural network for synthetic lethality prediction: overview, challenges, and future directions
    Briefings in bioinformaticsPeer-reviewedCC BY-NCFull textMolecular biology & gene regulationEurope PMC
  5. Pep2Net: an integrated peptide knowledge base and cross-attention model for multi-task prediction of peptides
    Boyan Zheng, Jiahao Yuan, Yuzhou Wang +4
    Genome BiologyPeer-reviewedCC BY-NC-ND 4.0Molecular biology & gene regulationCrossref

Cite

BibTeXRIS
BibTeX
@article{li2026katalyst,
  title = {{Katalyst: knowledge-guided semantic alignment via contrastive learning for enzyme turnover prediction}},
  author = {Ruilin Li and Senyu Tang and Jiaqi Deng and Taixing Qiu and Fei Guo and Jijun Tang and Xiaoyi Liu},
  journal = {Bioinformatics},
  year = {2026},
  month = oct,
  doi = {10.1093/bioinformatics/btag709},
  url = {https://academic.oup.com/bioinformatics/article/doi/10.1093/bioinformatics/btag709/8860687},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Katalyst: knowledge-guided semantic alignment via contrastive learning for enzyme turnover prediction
AU  - Ruilin Li
AU  - Senyu Tang
AU  - Jiaqi Deng
AU  - Taixing Qiu
AU  - Fei Guo
AU  - Jijun Tang
AU  - Xiaoyi Liu
PY  - 2026
DA  - 2026/10/03
JO  - Bioinformatics
DO  - 10.1093/bioinformatics/btag709
UR  - https://academic.oup.com/bioinformatics/article/doi/10.1093/bioinformatics/btag709/8860687
AB  - Motivation: The turnover number (kcat) is crucial for understanding enzyme kinetics, yet its experimental determination remains costly and low-throughput. Existing deep learning approaches often rely on single-substrate assumptions and overlook full reaction contexts or language-informed semantic representations, limiting their accuracy and generalization across diverse reactions.  Results: To address these challenges, we propose Katalyst, a knowledge-guided semantic alignment framework for kcat prediction based on contrastive learning. Katalyst integrates enzyme sequences, protein textual descriptions, substrates and products into a shared latent space using multi-channel encoders and contrastive alignment. An interaction-aware fusion module is designed to capture complementary cross-modal information. Experimental results demonstrate that Katalyst achieved the best overall performance, with the RMSE of 0.9880, R2 of 0.5829, and SRCC of 0.7603. Compared with the State-Of-The-Art (SOTA) baselines, Katalyst reduced RMSE by 3.01% and improved R2 and SRCC by 4.72% and 2.31%, respectively. Moreover, ablation studies and evaluations across sequence scales further confirmed the effectiveness of semantic alignment and multimodal feature fusion. Overall, Katalyst provides a robust and accurate framework for kcat prediction.  Availability and implementation: The code and data are available at https://github.com/liruilin-SUAT/Katalyst.
SN  - 1367-4803
SN  - 1367-4811
ER  -