Polypharmacology

De novo molecular design and in silico prediction of "polypharmacological" profiles are emerging research topics that will profoundly affect the future of drug discovery and chemical biology. The goal is to identify the macromolecular targets of new chemical agents. We develop and apply computational techniques, including "deep learning" and "big data" methods, for this purpose.

Computer-assisted decision making has found its place in modern medicinal chemistry. While in its comparably modest beginnings, the interactions of select few molecules with individual macromolecular receptors were almost exclusively studied by structure-based computational approaches. Today, high-quality experimental data amasses and allow us to address more complex questions involving multiple ligands, multiple binding sites and multiple receptor molecules. A ligand might interact with many targets, and a target may accommodate different types of ligands. "Polypharmacology" and the design of innovative drugs that affect a whole panel of targets in a predisposed manner have been recognized as a central issue in drug discovery. Furthermore, allosteric regulation by designer molecules has become tangible. For many diseases it may, therefore, no longer be sensible to pursue a one target–one drug philosophy. At the same time, the computer sciences have contributed fast hardware and computing solutions, as well as excellent algorithms that have already partially been transferred to the area of molecular informatics – in particular sophisticated machine-learning techniques for pattern recognition in large data sets and modeling of functional relationships between data classes.

Innovative computational tools can help in the analysis of the wealth of data constantly pouring in from newfangled biophysical, biochemical and biological measurements of compound effects on macromolecular targets, cells and tissues. Raw data from phenotypic screening and chemogenomics studies need to be analyzed, interpreted and used as a basis for hypothesis generation. This all points to personalized healthcare in the future, enabled by technological advancements in the generation and analysis of multi-dimensional data, as well as the rational design of new chemical entities with the desired activity spectra. Evidently, medicinal chemistry can benefit massively from the tight integration of modern computational concepts into drug-discovery processes.

JavaScript has been disabled in your browser