Padhi Siladitya, Ramakrishna Siddabattula, Priyakumar U Deva
Centre for Computational Natural Sciences and Bioinformatics, International Institute of Information Technology, Hyderabad, 500032, India.
J Comput Chem. 2015 Mar 30;36(8):539-52. doi: 10.1002/jcc.23828. Epub 2015 Jan 6.
An understanding of structure-function relationships of membrane proteins continues to be a challenging problem, owing to the difficulty in obtaining their structures experimentally. This study suggests a method for modeling membrane protein structures that can be used to generate a reliable initial conformation prior to the use of other approaches for sampling conformations. It involves optimizing the orientation of hydrophilic residues so as to minimize unfavorable contacts with the hydrophobic tails of the lipid bilayer. Starting with the optimized initial conformation for three different proteins modeled based on this method, two independent approaches have been used for sampling the conformational space of the proteins. Both approaches are able to predict structures reasonably close to experimental structures, indicating that the initial structure enables the sampling of conformations that are close to the native structure. Possible improvements in the method for making it broadly applicable to helical membrane proteins are discussed.
由于通过实验获得膜蛋白结构存在困难,因此理解膜蛋白的结构 - 功能关系仍然是一个具有挑战性的问题。本研究提出了一种膜蛋白结构建模方法,该方法可用于在使用其他构象采样方法之前生成可靠的初始构象。它涉及优化亲水残基的方向,以尽量减少与脂质双层疏水尾部的不利接触。基于此方法对三种不同的蛋白质进行建模,从优化的初始构象开始,使用了两种独立的方法对蛋白质的构象空间进行采样。两种方法都能够预测出与实验结构相当接近的结构,这表明初始结构能够对接近天然结构的构象进行采样。文中还讨论了使该方法更广泛适用于螺旋膜蛋白的可能改进措施。