论文部分内容阅读
全球20种最畅销的12种药物都是以GPCR(G蛋白偶联受体)为作用靶标,因为生化实验方法很难得到其三维结构,已知结构的GPCR数目只有7个.当前主流的GPCR结构预测方法主要是基于模板的预测和从头预测,但是由于GPCR序列相似性较低,而且长度普遍较长,故单靠一种方法很难预测出比较好的GPCR三维结构.本文结合GPCR的特点,将上述两种方法进行结合,提出了一种并行化的结构预测方法.使用该方法对GPCR Dock 2010中的两个目标(CXCR4和D3)进行预测,实验结果表明,本文预测的跨膜螺旋区域和ECL2区域比官网发布的大多数研究小组的结果更接近天然结构.
The world’s top 20 best-selling 12 drugs target GPCRs (G protein-coupled receptors) because of the difficulty of obtaining their three-dimensional structure in biochemical experiments, with only seven GPCRs of known structure. Current mainstream GPCRs Structure prediction methods are mainly based on template prediction and ab initio prediction, but because of the low similarity and long length of GPCR sequences, it is difficult to predict the better three-dimensional structure of GPCR by one method.This paper combines the characteristics of GPCR , A combination of the above two methods is proposed to predict the structure of two parallel targets (CXCR4 and D3) in GPCR Dock 2010. The experimental results show that the predicted transmembrane helix The results of the regional and ECL2 regions are closer to the natural structure than most research groups released by the official website.