Leveraging Large-Scale Functional & Developability Data to Generate Novel Antibody GPCR Agonists

  • Identifying GPCR antibody agonists from millions of NGS reads by learning sequence-activity maps with fine-tuned protein large language models
  • Integrating internal and public datasets to evaluate antibody developability to go beyond activity and increase the probability of success
  • Designing novel antibodies with high activity and superior developability by leveraging generative models to manage key trade-offs