Prediction and Analysis for Structure and Function of Membrane Proteins

该软件系统用于预测膜蛋白、信号肽和线粒体定位信号，并从全基因组中分析氨基酸序列。用于分析和预测的参数是多肽的电荷、疏水性、两亲性和由于甘氨酸和脯氨酸的密度而产生的柔韧性。所开发的系统用于(1)膜预测的改进，(2)信号肽的预测和(3)蛋白质的线粒体定位。利用膜蛋白预测的方法，分析了总基因组的所有氨基酸序列，发现膜蛋白的比例几乎是恒定的。从膜蛋白的进化世代角度讨论了跨膜螺旋的分布。采用量子化学、分子动力学和布朗动力学相结合的方法，对光受体和通道蛋白的动态结构和功能进行了模拟。

The software systems were developed for predicting membrane proteins, signal peptides and mitochondria localization signals, analyzing amino acid sequences from total genomes. The parameters for the analysis and prediction are the electric charges, the hydrophobicity, amphiphilicity and the flexibility of polypeptides due to the density of glycine and proline. The developed systems are for (1) the improvement of membrane prediction, (2) the prediction of signal peptides and (3) the mitochondria localization of proteins. Using the methods for the membrane proteins prediction, all amino acid sequences from total genomes were analyzed and the ratio of membrane proteins was found to be almost constant. The distribution of transmembrane helices was discussed in view of the evolutionary generation of membrane proteins. The simulations of the dynamical structures and the function of light receptors and channel proteins were performed by the combination of the quantum chemistry and the molecular dynamics as well as the Brownian dynamics.