Molecular topology analysis of the differences between drugs, clinical candidate compounds, and bioactive molecules

A new method to decompose molecules is proposed and used to analyze drugs, clinical candidate compounds and bioactive molecules. The method classifies a set of molecules into a few well-defined classes based on their molecular framework. It is then possible to use these classes to investigate differences between drugs, clinical candidates and bioactive molecules. The analysis shows that in comparison with clinical candidates and bioactive compounds, drugs have a higher fraction of compounds with only one ring system. This conclusion is still valid after correcting for lipophilicity (ClogP) and molecular size, as well as any potential protein target bias in the data sets. Furthermore the molecular bridge part of compounds in the drug set has on average fewer ring systems than molecules from the other sets. The ring system complexity (RSC) was also investigated and for most topological classes drugs have a lower RSC than the clinical candidates and bioactive molecules. Hence, this study highlights differences in topology between drugs, clinical candidate compounds and bioactive molecules.