Related Experiment Video
Updated: Jul 22, 2026
12:55
Scalable High Throughput Selection From Phage-displayed Synthetic Antibody Libraries
Published on: January 17, 2015
18.5K
An antibody developability triaging pipeline exploiting protein language models
James Sweet-Jones1, Andrew C R Martin1
1Institute of Structural and Molecular Biology, Division of Biosciences, University College London, London, UK.
Mabs
|March 5, 2025
Summary
This study introduces a machine learning pipeline to identify therapeutic monoclonal antibodies (mAbs) with favorable developability features. The approach uses protein language models to select promising antibody candidates early in development, reducing costs and improving success rates.
Area of Science:
- Biotechnology
- Immunology
- Computational Biology
Background:
- Therapeutic monoclonal antibodies (mAbs) are a key class of biologic drugs.
- While binding affinity is crucial, antibody sequence and structure significantly impact developability and clinical success.
- Late-stage failures due to poor developability are common and costly.
Purpose of the Study:
- To develop a computational pipeline for selecting developable antibody candidates from large libraries.
- To leverage machine learning and protein language models for early-stage antibody screening.
- To reduce the cost and improve the efficiency of therapeutic antibody development.
Main Methods:
- Utilized paired human antibody sequence data.
- Developed a machine learning pipeline employing protein language models.
- Clustered antibody sequences based on similarity to antibodies with known clinical success.
Main Results:
- Identified a method to predict antibody developability based on sequence features.
- The pipeline successfully identifies antibodies likely to possess favorable developability characteristics.
- Demonstrated a strategy to triage out antibodies with poor developability potential early.
Conclusions:
- The proposed pipeline is a valuable tool for selecting therapeutic monoclonal antibody candidates.
- This approach can significantly reduce the cost and time associated with antibody discovery.
- Facilitates the pursuit of novel therapeutics by improving early-stage candidate selection.
