Speaker: Yuyang Li, Graduate Student, Department of Statistics, Iowa State University

Title: Machine Learning Prediction Of Antibody Aggregation And Viscosity

Abstract:  In this presentation, I would like to share my intern experience in AstraZeneca this summer. The developability properties of monoclonal antibodies (mAbs), such as low aggregation propensity and low viscosity, are essential to new drug development. However, the stability profiles of antibodies at high concentrations are difficult to assess during early-stage discovery and candidate screening due to the limited number of molecules for which sequence, biophysical property data, and sufficient material are available. Therefore, development of predictive tools that can evaluate the developability of high concentration antibody formulation as early as possible in the discovery/development process is desired. I will introduce the background and several machine-learning related techniques in predicting antibody properties.