A general temperature-guided language model to design proteins of enhanced stability and activity
Designing protein mutants with both high stability and activity is a critical yet challenging task in protein engineering. Here, we introduce PRIME, a deep learning model, which can suggest protein mutants with improved stability and activity without any prior experimental mutagenesis data for the specified protein. Leveraging temperature-aware language modeling, PRIME demonstrated superior predictive ability compared to current state-of-the-art models on the public mutagenesis dataset across 283 protein assays. Furthermore, we validated PRIME's predictions on five proteins, examining the impact of the top 30 to 45 single-site mutations on various protein properties, including thermal stability, antigen-antibody binding affinity, and the ability to polymerize nonnatural nucleic acid or resilience to extreme alkaline conditions. More than 30% of PRIME-recommended mutants exhibited superior performance compared to their premutation counterparts across all proteins and desired properties. We developed an efficient and effective method based on PRIME to rapidly obtain multisite mutants with enhanced activity and stability. Hence, PRIME demonstrates broad applicability in protein engineering.
Authors
- Liang Hong (ORCID: https://orcid.org/0000-0003-0107-336X)
- Yi Xiong (ORCID: https://orcid.org/0000-0003-2910-6725)
- Pan Tan (ORCID: https://orcid.org/0000-0003-2086-0940)
- Fan Jiang (ORCID: https://orcid.org/0000-0002-9727-853X)
- Guisheng Fan (ORCID: https://orcid.org/0000-0002-2702-0242)
- Mingchen Li (ORCID: https://orcid.org/0000-0001-6862-0052)
- Liqi Kang (ORCID: https://orcid.org/0009-0004-2427-5807)
- Xinyu Sun (ORCID: https://orcid.org/0000-0002-2741-8963)
- Guangyu Yang (ORCID: https://orcid.org/0000-0002-2758-4312)
- Banghao Wu
- Yufeng Pei (ORCID: https://orcid.org/0000-0002-7515-6793)
- Lirong Zheng (ORCID: https://orcid.org/0000-0001-6803-5048)
- Shaojie Wang (ORCID: https://orcid.org/0000-0001-8278-3894)
- Jiajun Dong (ORCID: https://orcid.org/0000-0003-3471-5752)
- Wanli Ouyang (ORCID: https://orcid.org/0000-0002-9163-2761)
- Qian Liu (ORCID: https://orcid.org/0000-0002-6235-9065)
- Jia Liu (ORCID: https://orcid.org/0000-0001-9787-465X)
- Liang Zhang (ORCID: https://orcid.org/0000-0002-7672-1168)
- Xin Jin (ORCID: https://orcid.org/0000-0003-2211-2006)
- Jie Song (ORCID: https://orcid.org/0000-0002-4711-6014)
- Yan Feng (ORCID: https://orcid.org/0000-0002-2522-2115)
- Zhiqiang Hu (ORCID: https://orcid.org/0000-0003-0830-815X)
- Jin Huang (ORCID: https://orcid.org/0009-0003-9520-0610)
- Yuanxi Yu (ORCID: https://orcid.org/0000-0001-8682-5529)
- Yang Tan (ORCID: https://orcid.org/0009-0004-7261-1705)
- Wenxue Xu
Institutions
- University of Science and Technology of China (CN)
- East China University of Science and Technology (CN)
- Shanghai Jiao Tong University (CN)
- Chinese Academy of Sciences (CN)
- ShanghaiTech University (CN)
- Guangzhou Experimental Station (CN)
- Wuhan Institute of Bioengineering (CN)
- Beijing Academy of Artificial Intelligence (CN)
- Shanghai Advanced Research Institute (CN)
- Shanghai Institute of Applied Physics (CN)
- Group Sense (China) (CN)
- Institute of Natural Science (KP)
- Shanghai Artificial Intelligence Laboratory (CN)
Publication Details
- Journal
- Science Advances
- Published
- 2024-11-27
- DOI
- https://doi.org/10.1126/sciadv.adr2641
- Citations
- 39
- Primary Topic
- Protein Structure and Dynamics
- Type
- article
- Field-Weighted Citation Impact
- 9.12
Funders
- National Natural Science Foundation of China
- Shanghai Municipal Education Commission
- Shanghai Jiao Tong University
- National Key Research and Development Program of China