文摘
We present a novel computer algorithm, called GLARE (Global Library Assessment of REagents), thataddresses the issue of optimal reagent selection in combinatorial library design. This program reduces oreliminates the time a medicinal chemist spends examining reagents which a priori cannot be part of a "good"library. Our approach takes the large reagent sets returned by standard chemical database queries and producesoften considerably reduced reagent sets that are well-behaved with respect to a specific template. The pruningenforces "goodness" constraints such as the Lipinski rule of five on the product properties such that anyreagent selection from the resulting sets produces only "good" products. The algorithm we implementedhas three important features: (i) As opposed to genetic algorithms or other stochastic algorithms, GLAREuses a deterministic greedy procedure that smoothly filters out nonviable reagents. (ii) The pruning methodcan be biased to produce reagent sets with a balanced size, conserving proportionally more reagents insmaller sets. (iii) For very large combinatorial libraries, a partitioning scheme allows libraries as large as1012 to be evaluated in 0.25 s on an IBM AMD Opteron processor. This algorithm is validated on a diverseset of 12 libraries. The results that we obtained show an excellent compliance to the product propertyrequirements and very fast timings.