BRIEF
Accurate Prediction of Antibody Deamidations by combining high throughput automated peptide mapping and pre-trained protein language model-based deep learning
WATCH NOW

Discover groundbreaking advancements in the characterization of biologics with Ben Niu from BMS. His poster delves into the complex challenges of analyzing biologics, focusing on sequence motifs, deamidation, oxidation, and isomerization, and their impact on product stability, potency, and safety. Ben highlights the significance of efficient data generation and innovative machine learning models to predict sequence liability and hotspots, bypassing the need for labor-intensive peptide mapping. This approach not only streamlines the research process but also enhances high throughput screening, making it essential for anyone in early-stage biopharmaceutical development, including those working with fusion proteins.

His poster showcases cutting-edge methodologies, including a fully automated sample preparation platform that reduces months of work to mere days. Learn about advanced machine learning models that predict deamidation hotspots using protein language models and deep neural networks, leveraging both global and local sequence information for highly accurate predictions. Ben demonstrates how to use the information generated in our software for machine learning to quickly triage antibody clones, significantly cutting down research time from months to minutes. Visit the poster to explore these innovative techniques and discover how they can transform your approach to biologics characterization and protein engineering.

(Visit the paper pre-print)

SEE THE POSTER