Uniform coverage designs for molecule selection

In screening for drug discovery, chemists often select a large subset of molecules from a very large database (e.g., select 1,000 molecules from 100,000). To generate diverse leads for drug optimization, highly active compounds in several structurally different chemical classes are sought. Molecules can be characterized by numerical descriptors, and the chosen subset should cover the descriptor space or subspaces formed by several descriptors. We propose a method that concentrates on low-dimensional subspaces, a criterion for uniformity of coverage, and a fast exchange algorithm to optimize the criterion. These methods are illustrated by using a National Cancer Institute database.