Related Experiment Video
Updated: Jan 16, 2026
06:50
Author Spotlight: A Computational Approach to Decipher Amino Acid Preferences in Multispecific Protein-Protein Interactions
Published on: January 26, 2024
2.5K
Peptide design through binding interface mimicry with PepMimic
Xiangzhe Kong1,2, Rui Jiao1,2, Haowei Lin3,4
1Department of Computer Science and Technology, Tsinghua University, Beijing, China.
Nature Biomedical Engineering
|October 1, 2025
Summary
An AI algorithm, PepMimic, creates peptide binders for targeted therapy by mimicking binding interfaces. This approach yields high-affinity peptides, outperforming random screening and showing potential for diagnostics and therapeutics.
Area of Science:
- Biotechnology
- Artificial Intelligence
- Drug Discovery
Background:
- Peptides offer advantages for targeted therapy, including oral bioavailability, cellular permeability, and high specificity.
- Conventional small molecules and biologics have limitations in targeted therapy.
- Developing novel peptide binders is crucial for advancing targeted therapies.
Purpose of the Study:
- To develop an artificial intelligence (AI) algorithm, PepMimic, for designing peptide binders.
- To mimic binding interfaces of known targets and binders to create short peptide binders.
- To explore the potential of AI-designed peptides for diagnostic imaging and targeted therapeutics.
Main Methods:
- Developed PepMimic, an AI algorithm to design peptide binders by mimicking binding interfaces.
- Applied PepMimic to drug targets including PD-L1, CD38, BCMA, HER2, and CD4.
- Validated peptide binders using surface plasmon resonance imaging and in vivo mouse models.
Main Results:
- PepMimic successfully designed peptide binders with dissociation constant (KD) values as low as 10^-9 M.
- AI-designed peptides demonstrated significantly higher binding affinity compared to random library screening.
- Extensive validation in breast, myeloma, and lung tumor mouse models confirmed effective membrane binding.
Conclusions:
- PepMimic is a powerful AI tool for designing high-affinity peptide binders.
- AI-generated peptides show significant potential for clinical diagnostic imaging and targeted therapeutic applications.
- This approach advances peptide-based drug discovery and development.
