Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK.
Curr Opin Struct Biol. 2017 Aug;45:133-141. doi: 10.1016/j.sbi.2017.04.004. Epub 2017 May 13.
In this review, I discuss the recent advances in computational approaches to studying membrane protein structures, covering the latest methods for predicting a protein structure from its amino acid sequence, through to methods for assessing the structural dynamics and lipid interactions within molecular simulations of complex biological membranes. These approaches have not only benefited from advances in the computational software and architectures, but have also been assisted by a prodigious rise in the number of both the molecular sequences and experimentally determined membrane protein structures. The former, in part stimulated by metagenomics sequencing techniques, has led to an increased prediction accuracy for the computationally folded protein structures. The latter, assisted by improvements in structural biology approaches, has led to longer, larger and more complex molecular simulations of membrane proteins; many of which have greater relevance to human disease. Here I describe the methods for predicting a membrane protein structure from sequence, discuss the approaches to configure membrane protein simulations and detail the techniques used to identify and characterize specific lipid binding sites to membrane protein structures.
在这篇综述中，我讨论了计算方法在研究膜蛋白结构方面的最新进展，涵盖了从根据氨基酸序列预测蛋白质结构的最新方法，到评估复杂生物膜分子模拟中结构动力学和脂质相互作用的方法。这些方法不仅受益于计算软件和架构的进步，而且还得益于分子序列和实验确定的膜蛋白结构数量的大量增加。前者在一定程度上受到宏基因组测序技术的刺激，导致计算折叠蛋白结构的预测准确性提高。后者在结构生物学方法的改进的辅助下，导致对膜蛋白的更长、更大和更复杂的分子模拟；其中许多与人类疾病更相关。在这里，我描述了从序列预测膜蛋白结构的方法，讨论了配置膜蛋白模拟的方法，并详细介绍了用于鉴定和表征膜蛋白结构特定脂质结合位点的技术。
