Stocks Michael J, Wilden Gareth R H, Pairaudeau Garry, Perry Matthew W D, Steele John, Stonehouse Jeffrey P
Department of Medicinal Chemistry, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leics, LE11 5RH, UK.
ChemMedChem. 2009 May;4(5):800-8. doi: 10.1002/cmdc.200900050.
A practical and pragmatic method is demonstrated that aligns lead-like properties with compound diversity for the picking of compounds to synthesise from large virtual libraries. Methods are highlighted for decreasing synthetic attrition through the prior filtration of reagents sets grouped by reaction type. Also disclosed are protocols that use a combination of predicted physicochemical parameters and potential toxicological liabilities to enable the synthesis of lead-like compounds with a low potential risk of exhibiting toxicity or undesirable physicochemical properties. Lastly, a compound-picking process for a 2D compound matrix is demonstrated that maximises the diversity coverage whilst minimising synthetic effort. Thus a very highly optimised process is shown that delivers premium sample quality where lead-likeness and novelty are aligned to afford the best possible enhancement for the corporate compound collection.
展示了一种实用且务实的方法,该方法将类先导物性质与化合物多样性相结合,用于从大型虚拟库中挑选要合成的化合物。重点介绍了通过按反应类型对试剂集进行预先过滤来减少合成损耗的方法。还公开了一些方案,这些方案使用预测的物理化学参数和潜在毒理学风险的组合,以合成具有低毒性或不良物理化学性质潜在风险的类先导化合物。最后,展示了一种用于二维化合物矩阵的化合物挑选过程,该过程在最大程度减少合成工作量的同时,最大化了多样性覆盖。因此,展示了一个高度优化的过程,该过程提供了优质的样品质量,其中类先导性和新颖性相结合,为公司化合物库提供了尽可能好的增强。