Exploring Chemical Rings in a Simple Topological-Descriptor Space

A new method for organizing chemical rings based on their topology is presented. It uses three simple descriptors that characterize separate aspects of ring topology. These descriptors are integers and can thus be interpreted as the coordinates of discrete cells in a three-dimensional space. The descriptor values of any ring topology correspond to the coordinates of some cell. A database of rings can be distributed in this descriptor space by assigning each of them to the corresponding cell. This approach is applied to a database of 40 182 different ring topologies, derived from a comprehensive collection of chemical rings extracted from the CAS Registry File. This database is distributed among 7387 cells, and the population statistics and spatial distribution of these cells are discussed. An examination of selected cells shows that ring topologies which are similar tend to be close together in descriptor space. Some results of using this space to study ring diversity are presented. It is found that the distribution of the ring-topology database is not highly compact but has many significant voids. It is also found that the distribution of medicinally relevant rings in this space shows the influence of certain structural constraints on drug molecules.