Mechanistic Set Information
The mechanistic diversity set, which consists of 879 compounds, was derived from the 37,836 open compounds that have been tested in the NCI human tumor 60 cell line screen. In contrast to the original diversity set of 1,990 compounds, which was chosen on the basis of structural diversity, this mechanistic diversity set was chosen to represent a broad range of growth inhibition patterns in the 60 cell line screen, based on the GI50 activity of the compounds. Compounds that have been tested in the 60 cell line screen were clustered using the FASTCLUS procedure in the SAS statistical package. This algorithm is based on MacQueen's k-means algorithm, which minimizes the sum of squared distances from the cluster means. The procedure resulted in 1272 clusters. A single representative compound from each cluster, for which an adequate supply of material was available, was chosen. Some clusters are not represented in the set, as insufficient material was available.