From Origin to Current Methods: An Overview of Molecular Modeling Applied to Medicinal Chemistry in the Last 30 Years

Molecular modeling soon attracted the attention of Medicinal Chemistry researchers, given the importance of molecular structure for understanding the mode of action and for designing bioactive compounds. Computer-assisted drug design (CADD) has become widespread and today big pharmaceutical companies routinely uses it as a support tool in the search for new drugs. Here, it will be addressed the most relevant topics about CADD from the last 30 years, when research groups in Medicinal Chemistry began to explore molecular modeling in Brazil. This history can be described through phases in which some methods emerged and became predominant, in a continuous evolution, passing, for example, from the basic empirical field and quantum mechanics molecular modeling procedures, through quantitative structure-activity relationship (QSAR) methods, molecular docking and, more recently, virtual screening. Since the mid-2000s, machine learning methods have been increasingly applied to the solution of problems in the context of Medicinal Chemistry, such as the determination of protein 3D structure and the characterization of relationships between chemical structures and their biological activities. Far from being complete, this history continues its evolution, bringing significant contributions to the drug design, either by reducing the time and cost of research, or by enabling and accelerating the finding for new bioactive compounds.

Keywords:
molecular modeling; CADD; LBDD; SBDD; machine learning