基于虚拟活性谱的药物重定位研究

Drug repositioning is to identify new uses for existing drugs or compounds, which would reduce the time and costs of drug discovery and hence has become a matter of intense interest in drug development. Previous studies demonstrated that compounds with similar efficacy could be identified based on similar in vitro activity profile. Moreover, He et al. reported that similar results could be achieved based on in silico activity profile. Here, drug repositioning will be carried out based on the virtual dockings of 8,714 small molecule drugs from CMC (Comprehensive Medicinal Chemistry) against 7,998 clinical or potential drug targets. Through literature survey and data mining, target selection algorithm will be optimized and a key targets group will be identified for a specific class of drugs with similar efficacy. If other drugs have similar in silico activity profile against the key targets group, then bioassays will be performed in the hope that 3~5 repositioned drugs will be identified as anticancer or antidiabetic agents with 1~2 exhibiting activity in animal model. This project will establish the strategy for drug repositioning based on large scale virtual screening, which is of tremendous scientific and practical value to old drugs for new uses and to innovative medicine of our country.

药物重定位，即研究已有药物或者化合物原有适应症之外的新用途，可以大大缩短研发时间并减低研发成本，已经成为新药研发的一个重要方向。国际上的研究表明基于体外活性谱的相似性可以发现具有类似药效的化合物，国内贺林院士等的工作表明虚拟活性谱也有类似的效果。本项目拟针对综合药物化学数据库（CMC）中的8,714个小分子药物开展药物重定位研究。首先完成CMC中的药物分子对7,998个临床及潜在药物靶标的分子对接；在此基础上，结合文献检索和数据挖掘，优化潜在靶标的选取方法，获得具有相同药效的一类药物所对应的关键靶标群；然后对关键靶标群的虚拟活性谱进行比较分析，预测有类似药效的其他药物；最后进行药物重定位实验验证，发现3~5个有体外活性的抗肿瘤及糖尿病等疾病的药物，其中1~2个有动物活性。本项目建立的基于大规模虚拟筛选的药物重定位策略和技术对我国的老药新用及创新药物研究都具有重要的学术意义和应用价值。

药物重定位，即研究已有药物或者化合物原有适应症之外的新用途，它可以大大缩短研发时间并减低研发成本。前人研究发现，体外活性谱相似的化合物具有类似的药效。本课题针对综合药物化学数据库（CMC）中的8,714 个小分子药物开展药物重定位研究。将CMC 中的药物分子对7,998 个临床及潜在药物靶标的分子对接，得到虚拟活性谱；然后对虚拟活性谱进行比较分析，预测有类似药效的其他药物；最后进行药物重定位实验验证，发现了6个有体外活性的药物，其中针对黑色素瘤的重定位药物2个（雷复尼特、氯生太尔）、抗黑色素生成的重定位药物2个（petasiphenol、柳胺酚）、抗骨髓瘤/淋巴瘤的重定位药物2个（紫檀芪、多巴酚丁胺）；其中紫檀芪具有动物活性。在小鼠淋巴瘤移植瘤和骨髓瘤移植瘤模型上可以显著抑制肿瘤的生长。此外，开发了一天可以虚拟筛完4200万化合物的分子对接软件mD3DOCKxb以及发现卤键调节药物的ADME/T性质。发表SCI 论文5篇，申请专利3 件，完成了预期的研究目标。