Computer-aided ligand discovery and hit optimization of small molecules targeting the TGR5 receptor, Dengue Virus protease, and Pim-1 kinase

In the last 50 years, computational drug discovery strategies have been implemented and optimized to accelerate lead compound identification. When supported by computational approaches, the time, costs, and labor spent on drug discovery exclusively with conventional methods can be significantly reduced. The scientific field of early-stage drug discovery is being revolutionized by the new potential offered through supercomputing power, artificial intelligence, and new technologies. As a consequence, the scientific community produces a large number of experimental results. Therefore, using computational methods to integrate, analyze, or visualize data is essential. This thesis approaches three distinct scientific problems in the pharmaceutical field, under the lens of computational methods. The aim is to modulate the biological activity of three distinct protein targets with small molecules using molecular simulations and modeling. More specifically, the first scientific challenge is the activation of the human transmembrane receptor TGR5, a target against metabolic and inflammatory diseases, by non-steroidal small molecules that act either as agonists or as positive allosteric modulators. The second scientific subject focuses on potential treatment against the disease caused by the Dengue virus, via allosteric inhibition of its protease. The last scientific argument involves inhibiting the Pim-1 kinase, a well-studied prooncogenic target, with the scope to optimize existing scaffolds. Despite the numerous differences among these proteins and the targeted diseases, the in silicon methods herein presented are based on analogous concepts and, for the most part, structure-based computational techniques. The calculations performed in this study rely on the three-dimensional structures of these targets that have been experimentally resolved. The main focus is to identify novel small molecules that modulate the biological activity of the proteins by binding to them and, therefore, determining the activation or inhibition of the protein target. For all three targets, virtual screening campaigns were performed employing molecular docking calculations to identify small molecules capable of inducing the desired biological response. The identification of the small molecules was performed by employing molecular dynamics simulation techniques and chemoinformatic methods –such as similarity searches– were used in this thesis.