The broader impact/commercial potential of this Small Business Innovation Research (SBIR) project is to provide a tool for designing new drugs. Designing new drugs that bind to a specified protein target required finding the best molecule in a vast chemical space. The proposed design tool can search this large chemical space much more efficiently and cheaply compared to current methods. It is based on a "reverse engineering" method to solve the problem of going from a set of desired properties back to chemical structures that may have these properties. This design tool allows the efficient search of the database to find the best molecule in advance of laboratory work. The envisioned technology is expected to shorten the screening and drug discovery phase from 3 years to 1 year, while saving ~$20 M per target overall, by shifting current practices and enabling a rapid, lower cost, and more novel targeted drug discovery and lead identification. This project proposes to develop a computational drug discovery platform. This platform uses advanced QM/MM (quantum chemistry and molecular chemistry combined) calculations for binding accuracy combined with an artificial intelligence heuristic search algorithm to find the most appropriate molecule in a vast molecular space. The advances to be accomplished with this SBIR project include adding two novel toolsl to improve the performance of the current "in silico" drug design algorithm: 1) a multi-object optimization algorithm to further improve high accuracy, and 2) an automated scaffold design algorithm. These will result in highly selective designs with novel scaffolds that not only bind well to the target, but that also have excellent drug-like properties such as low probability of toxicity and off target effects, as well as greater stability and synthesizability. These advancements should provide higher accuracy of binding affinity prediction results with fewer false positives and better molecule selection, and reduced labor, which allows for greater automation.
