Related Experiment Video
Updated: Aug 5, 2026
08:49
Incorporating Target Protein Structure Flexibility and Dynamics in Computational Drug Discovery Using Ensemble-Based Docking Analysis
Published on: June 20, 2025
Integrating AlphaFold2 with physics-based ensemble docking for high-efficiency nanobody discovery
Yinghao Guo1, Renfang Guan1, Lunde Jin1
1Department of Immunology, School of Basic Medical Sciences, Harbin Medical University Harbin 150081 China jma@hrbmu.edu.cn xingwenjing@hrbmu.edu.cn.
Chemical Science
|August 1, 2026
Summary
This study introduces a computational framework to efficiently discover functional nanobodies. The method uses AI to predict structures and binding energies, improving hit recovery and reducing redundancy in nanobody selection.
Area of Science:
- Biotechnology and Pharmaceutical Sciences
- Computational Biology and Bioinformatics
- Immunology and Molecular Biology
Background:
- Nanobodies, small single-domain antibodies, offer therapeutic potential but their discovery is hampered by inefficient experimental panning.
- Identifying functional nanobodies from large libraries requires overcoming bottlenecks in selection and validation processes.
Purpose of the Study:
- To develop and validate a computational framework for prioritizing functional nanobodies from next-generation sequencing (NGS) data.
- To improve the efficiency and reduce redundancy in nanobody discovery compared to traditional methods.
Main Methods:
- Utilized AlphaFold2 for predicting nanobody structures from NGS-derived sequences.
- Employed ensemble docking and MM/GBSA re-scoring for predicting nanobody-antigen binding poses and energies.
- Validated computational predictions using experimental methods like flow cytometry, ELISA, and surface plasmon resonance (SPR).
Main Results:
- The computational framework successfully prioritized candidate nanobodies against Mesothelin (MSLN), PD-1, and Nectin-4.
- Over 70% of top-ranked candidates showed strong binding, with nanomolar affinity confirmed for representative nanobodies.
- The workflow enhanced hit recovery and reduced clone redundancy compared to random selection methods.
Conclusions:
- The developed computational framework significantly improves the enrichment of functional nanobodies from NGS-derived pools.
- This approach offers a practical and efficient alternative to conventional panning methods for nanobody discovery.
- The study demonstrates the utility of integrating AI-driven structure prediction and binding energy calculations in antibody engineering.
