反向虚拟筛选是以活性小分子为探针、通过计算化学从蛋白质结构数据库中发现潜在作用靶标的新技术。海洋天然产物活性高但含量低，常常无法提供足够样品量进行常规靶标研究，靶标筛选成为海洋药物研究的瓶颈，反向虚拟筛选技术因而尤为重要。我们应用这一技术成功发现了新骨架h(p)300抑制剂swinhoeisterol A，并建立了计算机筛选模型。研究发现，现有模型的筛选结果与生物测试结果有时存在较大差距，致使这一技术的普遍应用受到限制。本课题拟对现有的h(P)300蛋白抑制剂筛选模型进行优化及方法学验证，运用新模型预测活性化合物不同功能位点对活性的可能影响，并通过分离及合成方法得到相应化合物对预测结果进行生物学验证，验证结果用于模型的反馈性改进及优化，最终建立较为可靠的h(P)300抑制剂计算机虚拟筛选模型，为新型h(P)300抑制剂的发现提供技术支持，同时为反向虚拟筛选方法学改进提供参考。

“Inverse virtual screening” (IVS) is a new technique of computational chemistry using bioactive molecule as probes to discovery potential targets from protein data bank. Marine natural products are known to have potential bioactivity but their contents are generally very low in marine organisms. It is therefore not surprise that the amounts of many compounds are not sufficient to match the request of target discovery using common biological approaches. Potential target screening of bioactive compounds are regarded as a choke point in the field of marine drug discovery. IVS is particularly useful in the target screening of marine bioactive molecules. The technique was successfully employed in my lab in discovery of the new class of h(p)300 inhibitor, swinhoeisterol A. A computational model was established simultaneously for screening of h(p)300 inhibitor. Our research indicated a fairly difference between the IVS result and that of biotest. This actually limits the widely application of the newly emerging technique. The current program intends to optimize and methodological validate the established screening model, using the optimized model to predict the possible biological affects of functional groups in bioactive molecules, and to validate their h(P)300 inhibitory activity using correlated molecules obtained by isolation and chemical synthesis. The validation will direct a further optimization of the model to establish a dependable computational model in screening of h(P)300 inhibitor. The research will give not only a technique support in discovery of new class of h(P)300 inhibitor but also a reference in improvement of IVS methodology.