Therapeutic monoclonal antibodies bind to specific regions of proteins called epitopes, which elicit cellular responses that treat or cure disease. Discovering therapeutic antibodies traditionally requires laborious and expensive screening experiments, so computational approaches that select which antibodies bind an epitope best and have the most desirable pharmaceutical properties are in high demand. Structure-based antibody design is also important to the modern drug discovery and development process. This approach requires a high- resolution quaternary (3D) protein complex structure, whose experimental determination is often a slow process that is not always successful. Protein structure and binding interface prediction algorithms are poised to impact human health by accelerating the construction of high-confidence structural models of drug targets and biopharmaceuticals, which will help identify new therapeutic strategies. However, the current algorithms are limited in their ability to distinguish stronger-binding antibodies from weaker ones, which is preventing the discovery of broad classes of therapeutics. In addition, technologies are needed to predict if a candidate antibody will fail as early as possible in the development process. With improvements in simulating removal of molecular liabilities without damaging function, computer-aided antibody design can be used to lower drug development costs and focus experiments on the most promising drug candidates. Here we propose to advance antibody discovery by developing highly accurate software tools built on the success of DNASTAR’s NovaFold Antibody program for antibody structure prediction, NovaDock for flexible protein-protein docking, and Lasergene Protein Design for protein engineering. The aims of the project focus 1) on developing more accurate and effective immune complex (an interacting antibody and antigen) structure predictions through better modeling of the challenging complementarity-determining regions (CDR), which play a critical role in antibody affinity and selectivity; and 2) on predicting antibody sequences that reduce chemical and energetic liabilities that prove detrimental to an antibody’s manufacturing process or therapeutic effect in a patient. In particular, overall predictive capability will be improved by incorporating computational acceleration techniques to support the virtual screening of tens of thousands of antibody sequences. Finally, and for the first time, this project will develop a “virtual immune system” to approach human antibody discovery, where antibodies will be modeled from germline sequences and selected for best recognizing an antigen of interest. The overall project goal is to deliver an advanced antibody screening pipeline that is powerful, accurate, and produces fast results, which will accelerate antibody discovery by enabling detailed and accurate immune complex structure predictions and structure-based liability detection at a high-throughp...