| Abstract
| - Compound selection methods currently available to chemists arebased on maximum or minimum dissimilarityselection or on hierarchical clustering. OptimizableK-Dissimilarity Selection (OptiSim) is a novelandefficient stochastic selection algorithm which includes maximum andminimum dissimilarity-based selectionas special cases. By adjusting the subsample size parameterK, it is possible to adjust the balance betweenrepresentativeness and diversity in the compounds selected. TheOptiSim algorithm is described, alongwith some analytical tools for comparing it to other selection methods.Such comparisons indicate thatOptiSim can mimic the representativeness of selections based onhierarchical clustering and, at least insome cases, improve upon them.
|
