Quantitative Structure‐Activity Relationships and Computational Methods in Drug Discovery

Abstract

Among the various methods for finding new medicinal drugs, the approach using molecular descriptors for quantitative structure–activity relationships (QSAR) is described in more detail, after a brief historical introduction. Starting with Corwin Hansch's observation that for many organic compounds their biological activities can often be correlated with their lipophilicity (octanol–water partition coefficients), which in turn may be approximated by computational approaches, chemical structures proved to be the early key to drug design. How can one encode discrete chemical structures into numerical values that may then be compared with biological activities that are measured by continuous scales? One well‐working tool uses topological indices (TIs) that are graph‐theoretical invariants associated with the hydrogen‐depleted constitutional formulas. Several examples are provided evidencing that this pathway proved to be successful. For three‐dimensional (3D) chemical structures, more sophisticated theoretical methods (such as CoMFA) had to be implemented. Nowadays, powerful synthetic and analytical methods involving combinatorial chemistry and high‐throughput screening on one hand, together with refined computational approaches, try to lower the cost of each new medicinal drug, which may reach one billion US dollars.