RATIONAL SCREENING SET DESIGN AND COMPOUND SELECTION - CASCADED CLUSTERING

The use of cascaded clustering is reported. This technique was develop ed to permit the application of Jarvis-Patrick clustering based on str uctural fingerprints to large chemical databases, while keeping the ma ximum cluster size and the number of singletons produced at reasonable levels. The basis for the algorithm, its implementation, and validati on are described. in the first part of the paper, the approach is used to create a representative subset of compounds for biological testing from the corporate compound repository. A variation of the method is then used for the comparison of relatively large databases. Finally, c ompound selection using cascaded clustering is shown to be complementa ry to the Diverse Property Derived approach, which is based on partiti oning by six molecular descriptors.