Molecular modeling and design of new ligands of biological interests

Correct 3D representation (conformation) of chemical structures is fundamental to understanding the physico-chemical and biological properties associated with these structures. Several mathematical models are available to define the conformations associated with each structure, from quantum mechanical to molecular mechanical methods, taking into account the complexity of the molecule and the properties we wish to model. These methods enable us to quickly obtain a first probable conformation, taking into account the limited flexibility of these structures. However, the conformational flexibility of molecules quickly becomes a major problem to integrate, even for simple molecules, through the notion of conformational space. The first part of this article recalls a number of basic elements concerning molecular mechanics and the treatment of molecular flexibility.

The historical approach to ligand design is associated with chemogenomics, which considers that a specific interaction between a molecule and a biological receptor (affinity/inhibition) leads directly to an observable phenotype (cellular, tissue, general). This logic is taken up by therapeutic chemistry, with the characterization of compounds presenting an initial affinity followed by optimization of their affinity and selectivity through structural modulations. Molecular modeling techniques are at the heart of this approach, with a three-dimensional analysis of interactions between single molecules and macromolecules, enabling us to understand and anticipate a biological response to a receptor. The aim is to analyze these interactions in order to understand the affinity of derivatives, and then to characterize new chemical families or help optimize the first ligands. As a result, molecular modeling describes intermolecular interactions to guide rational ligand design. In this article, two approaches are presented: one based on the comparative analysis of ligands and the other based on the analysis of ligand-receptor interactions.

In the 2000s, systems biology emerged with the integration of omics data in the form of models to understand an observed global phenotype. The positioning of molecular modeling has shifted, taking into account biological networks in order to better understand the phenotype, a finer integration of selectivity phenomena and the interpretation of results from systems biology.

The approaches described here are part of a broader discipline that has been developing in France over the last ten years or so: chemoinformatics.

The aim of this article is to provide the reader with an initial overview of the methodologies used to identify new ligands of biological interest using so-called in silico computational techniques. The characterization...