应用同源建模的蛋白质结构预测已经成为一种快速获得蛋白质结构的技术,这种技术也将成为完成结构基因组计划的有力工具。同源建模是指寻找与目标序列同源而且有实验测定结构的蛋白质作为模板,从而构建目标序列的结构模型的方法。限制这种方法的应用主要是同源建模的关键步骤,即目标与模板之间序列比对和环区建模的准确性。当模型的准确性达到令人信服的程度时,更为精确的计算机辅助药物设计和改造蛋白质,甚至设计全新功能的蛋白质将成为可能。综述了从算法和策略上提高同源建模关键步骤准确性的研究进展。
Protein structure prediction by using homology modeling has become a fast technique to obtain protein structure and offer an power tool to complete structure genomic. Homology modeling means search the protein which has structure determined by experiment and has homology relating with the target sequence for template, and then build the structure model of the target sequence. Limiting the usage of this technique major on the key procedures, which is the fidelity of target-template alignment and the loop modeling. When the accuracy of the model reached higher level, more exact to rebuild protein and computer aided drug design, even designed protein with whole new function, can be obtained possibly. Trends in the researches on improve the accuracy of target-template alignment and loop modeling by methods and algorithms were summarized.