Accelerating antibody discovery and optimization with high-throughput experimentation and machine learning

Area of Science:

• Biotechnology
• Computational Biology
• Immunology

Background:

• Antibody engineering is crucial for developing novel therapeutics.
• Traditional methods face challenges in efficiency and scope.
• Data-driven approaches are emerging to address these limitations.

Purpose of the Study:

• To review advancements in integrating high-throughput experimentation and machine learning for antibody engineering.
• To highlight how these methods enable rational design and optimization of antibody therapeutics.
• To discuss current challenges and future directions in AI-powered antibody discovery.

Main Methods:

• Utilizing extensive datasets of antibody sequences, structures, and functional properties.
• Training predictive machine learning models, including protein language models.
• Integrating experimental data acquisition with computational feature extraction.

Main Results:

• Machine learning models can predict and optimize antibody affinity, specificity, stability, viscosity, and manufacturability.
• Synergy between experimental and computational approaches accelerates antibody engineering.
• Demonstrated through practical examples and case studies.

Conclusions:

• AI and high-throughput experimentation significantly transform antibody discovery and optimization.
• Capturing both sequence and structural information is vital for predictive model accuracy.
• Further development is needed to fully realize AI's potential in expediting therapeutic development.