本文以17个PDE4B受体-抑制剂复合物为研究对象，应用LigandScout构建17个药效团模型，并以多药效团组合筛选方法开展中药中具有PDE4B 亚型选择性抑制活性成分的辨识研究。经过配体再匹配和已知活性化合物数据库筛选两种评价方法，评价并保留了17个药效团模型；以分子被命中频次作为多药效团组合筛选方法的综合性评价指标，结合MDDR 数据库筛选结果，确定了PDE4B抑制剂被命中大于或等于8 次的分子为多药效团组合筛选目标。比较了最优药效团筛选方法与多药效团组合筛选方法的筛选效率，结果多药效团组合筛选方法的富集度和活性化合物有效命中率都比最优药效团筛选方法高，分别为38.017、47.44%，表明基于受体的多药效团组合筛选方法具有一定的可靠性，有助于开展中药中有效成分的辨识研究。

Seventeen receptor-based pharmacophore models of phosphodiesterase 4B (PDE4B) inhibitors were generated by LigandScout based on seventeen PDE4B receptor-inhibitor complexes. The potential PDE4B selective inhibitors were identified from traditional Chinese medicine (TCM) with the method of combinatorial screening with seventeen models. By re-fitting with the ligand and screening MDL Drug Data Report (MDDR, Version 2007.2), seventeen pharmacophores were evaluated. Based on the screening results of MDDR, hit frequency of molecules was defined as the comprehensive appraisal index of combinatorial screening. And the molecules hit by not less than eight pharmacophores were taken as the screening objects of PDE4B inhibitor. The screening efficiency by the best pharmacophore and combination of multiple pharmacophores was compared. The result showed that the enrichment and percentage of actives in the hit list of combinatorial screening were better than the best pharmacophore screening, which was 38.017 and 47.44% , respectively. The results showed that combinatorial screening method of receptor-based multiple pharmacophores is reliable and contribute to identify the active compounds from TCM. Traditional Chinese Medicine Database (TCMD, Version 2005) was screened with combinatorial screening method of multiple pharmacophores and the results would give guidance for further study of PDE4B inhibitors.

科学技术部国家科技支撑计划（2008BAI51B01）：中药有效成分群辨识技术研究，负责人：张宏桂；国家自然基金项目（81173522）：点线面三维模拟活血化瘀中药防治冠心病的作用机制研究，负责人：张燕玲；教育部博士点基金项目（20092213120006）：基于花生四烯酸代谢途径的清热解毒中药有效成分信息辨识研究，负责人：张燕玲；北京中医药大学自主课题（2011-CXTD-11）：中药信息工程，负责人：乔延江；北京中医药大学自主课题（2011-JDJS-12）：活血化瘀中药治疗冠心病的作用机制研究，负责人：张燕玲。