以高通量(high throughput)且有效的方式找出並修改設計對人體具相當療效的多靶點抗癌中草藥衍生物，以ligand-based 和structure-based 兩種不同方式進行平行比對運算，並且在開發出新藥後與有機合成老師及生化老師合作共同完成新藥開發。本計畫先以實驗室既有的中草藥資料庫為出發點，與研究多年的主題COX-2、mPGES-1、Her2、 c-Met、Hsp90、Hsp70 以及AMPK等與癌症相關的蛋白質為藥物設計對象，依此再深入研究相關訊息傳導路徑、多重靶點藥物設計，且將在在計畫執行中繼續加入其他癌症相關蛋白質為研究對象。第一年先以蛋白質結構為基礎(structure-base)的研究方式以分子對接及藥物de novo 設計為主。針對不同種類蛋白質細分不同蛋白質家族，利用3D-QSAR 尋找相同特性區域，以統計方式尋求最佳的藥物分子。另一設計藥物之方式為de novo design 的策略，以蛋白質的活性位置為基礎尋找其胺基酸上的重要區域，以此對接有藥效的藥效基團，直接設計有高藥效藥物。第二年將以藥物結構為基礎的藥物篩選以及設計。利用資料庫中具有已知活性的大筆藥物資料，進行CoMFA、CoMSIA 的迴歸分析，利用分子疊合將最適切的結構特性模型尋出，並預測其藥物活性。而後將這兩種方法所做出的模擬結構，鍵入資料庫中，並配合原有資料庫中有效藥物的pharmacophore 進行測試及修改，ㄧ層層篩出最具效力的藥效基團。第三年在多靶點抑制藥物的設計上，會先將具有共同特性的蛋白質分類，依此在設計流程上將會比較容易著手亦有其合理性。以蛋白質為基礎的藥物對接與以藥物結構為基礎的共同特性尋找之兩大類方法尋找其共同之特性並加以修改，設計出一次就可抑制多個蛋白質的潛力藥物。第四年第五年 重複上述步驟，直到把先導藥物修飾到無毒性、水溶性佳、具有經濟的合成價值。完成癌症相關基因訊息傳遞路徑分析圖，完成癌症相關基因藥物資料庫。
This project focuses on the effectively multiple-targeted anticancer drugs from our laboratory chinese herbs database by computer-aided high throughput screening and drug design. The ligand-based and structure-based screening manners will employ first and follow drug synthesis and bioactivity test by assistance of the cooperator proceed. Our Chinese herbs databank about anticancer agents and well-researching targets including COX-2、mPGES-1、 Her2、c-Met、Hsp90、Hsp70 and AMPK serve firstly for drug design. The further investigation of relative signal pathway, multiple-targeted drug design and new target search are performed latter. At first year, the molecular docking and de novo drug design proceed basing on the protein structure (structure-based manner). The QSAR study revealing the relative properties of different protein active sites can develop the potent compounds with compatible features. Additionally, the progress of de novo design enhances the interaction affinity of effective compounds by substitution and addition of the functional group basing on the surrounding residues. The virtual screening and drug design basing on the compound structure (ligand-based manner) perform at second year. The bioactivity data of considerable drugs from database is utilized for regression analysis by CoMFA and CoMSIA. The molecular alignment describes the structural properties and the bioactivity is predicted consequently. The QSAR model is employed for improvement of the potent compounds by changing pharmacophore of ligands. The properties of influential pharmacophore are determined. According to the project of multiple-targeted drug design, the characteristics of diverse proteins are initially assorted for convenience and reasonableness. The two major manners by molecular docking (structure-based) and structural properties (ligand-based) search the comparatively common features and design the potent drugs against multiple targets. At the end of this project, we will accomplish the oncogene signal pathway, the oncogene-ligands database, the patent of multiple target ligands.
