Developing robust zero-shot AI models for anti-aging antibody design

About This Grant

Project Summary Anti-aging antibody research, including strategies targeting interleukins and other antigens, shows promise in rejuvenating the immune system, improving metabolic functions, and extending healthy lifespans. AI-driven platforms are revolutionizing antibody development by accelerating affinity maturation and optimizing developability properties, enabling simultaneous optimization of multiple characteristics. These advancements could lead to more effective treatments for age-related diseases and a significantly improved quality of life for the growing aging population. However, zero-shot predictions for antibody affinities using pretrained models without additional target-specific data remain challenging. In this project, we propose a new strategy to address this challenge by generating diverse antibody-antigen interactions at an unprecedented scale (Aim 1) and training new AI models using these generated data in combination with data collected from literature and public databases (Aim 2). We will rigorously evaluate the performance of the new models and benchmark against the state-of-the-art methods. We will test the generality of the new models on a diverse set of antigens and experimentally validate the prediction accuracy (Aim 3). We will apply the models to identify new antibodies against new therapeutic targets associated with ageing or age-related diseases. Once complete, the proposed research will provide a powerful tool for accelerating antibody discovery and optimization as well as new antibody candidates for anti-aging treament.