ABSTRACT
Virtual Screening (VS) methods can drastically accelerate global drug discovery processes. Among the most widely used VS approaches, Shape Similarity Methods compare in detail the global shape of a query molecule against a large database of potential drug compounds. Even so, the databases are so enormously large that, in order to save time, the current VS methods are not exhaustive, but they are mainly local optimizers that can easily be entrapped in local optima. It means that they discard promising compounds or yield erroneous signals. In this work, we propose the use of efficient global optimization techniques, as a way to increase the quality of the provided solutions. In particular, we introduce OptiPharm, which is a parameterizable metaheuristic that improves prediction accuracy and offers greater computational performance than WEGA, a Gaussian-based shape similarity method. OptiPharm includes mechanisms to balance between exploration and exploitation to quickly identify regions in the search space with high-quality solutions and avoid wasting time in non-promising areas. OptiPharm is available upon request via email.
