蛋白质亚细胞定位预测研究进展

蛋白质的功能与其在细胞中的定位有着密切的联系，新合成的蛋白质必须处于适当的亚细胞位置才能正确的行使其功能．预测蛋白质的亚细胞定位，在确定一个未知蛋白质的功能，了解蛋白质相互作用等方面有着重要的意义．机器学习方法在蛋白质亚细胞定位研究中扮演着一个重要的角色．笔者从数据集的构建、蛋白质序列特征提取方法、蛋白质亚细胞定位预测算法以及预测算法的性能评估等四方面总结了过去十几年间机器学习方法在蛋白质亚细胞定位研究中的应用情况，系统阐述了蛋白质亚细胞定位预测研究的进展．

ADVANCEMENT OF PREDICTING PROTEIN SUBCELLULAR LOCALIZATION SITES

Protein function localization in ceils are closely linked. Newly synthesized proteins must be in proper subcellular location to properly exercise their functions. Prediction of protein subcellular localization has important significance in the setting of an unknown protein functions and understanding of protein - protein interations. Machine learning Methods play an important role in protein subcellular localization research. This article summevrizes machine learning methods in the application of protein subcellular localization of the past decade in four aspects. It involves the data set, protein sequences feature extraction method, prediction of protein subcellular localization prediction algorithm and the algorithm performance evaluation. And it describes the advancement of predict protein subcelluar localization completely.