评价不同的打分函数对虚拟筛选β-分泌酶抑制剂富集率的影响

Effect of different scoring functions on enrichment factors of virtual screening for β-secretase inhibitors

目的研究多种打分函数对计算机虚拟筛选β-分泌酶(β-secretase,BACE-1)抑制剂的影响。方法采用分子对接软件Surflex虚拟筛选了50个BACE-1的抑制剂和9950个无活性分子,同时应用5种打分函数对筛选的结果分别应用单个打分函数和多个打分函数组合得分排序。结果单独以Surflex_score一个函数抽提对接后的结合模式再进行打分排序后富集率为40,将Surflex_score和D_score两个函数组合后排序可获得48的富集率。结论组合多种不同的打分函数比单个打分函数打分能获得更多的活性化合物。

Objective To explore the impact of the different scoring functions on the enrichment factors of virtual screening for β-secretase（BACE-1）inhibitors. Methods Virtual screening of the database composed of 50 BACE-1 inhibitors and 9950 inactive compounds for BACE-1 inhibitors were performed using docking program of Surflex, the results were scored and ranked with different single scoring functions and combine scoring functions. Results The enrichment factor（1%） was 40 when using the single surflex_score, the enrichment factor （1%） was 48 when combining the surflex_score and D_score to rank. Conclusion We prefer combine scoring functions to single scoring function because it can obtain more active compounds.