蛋白质的亚细胞定位是进行蛋白质功能研究的重要信息.蛋白质合成后被转运到特定的细胞器中,只有转运到正确的部位才能参与细胞的各种生命活动,有效地发挥功能.尝试了将保守序列及蛋白质相互作用数据的编码信息结合传统的氨基酸组成编码,采用支持向量机进行蛋白质亚细胞定位预测,在真核生物中5轮交叉验证精度达到91.8%,得到了显著的提高.
Subcellular localization is a key characteristic of protein functional research. Proteins are transported to specific compartment after they are synthesized in cells. They can take part in the cell activity and function efficiently when in correct subcellular location. Sequence homolog, protein-protein interaction information and traditional amino acid composition are combined as input parameters of support vector machine (SVM) to predict eukaryotic protein subcellular localization. The total accuracy of 5-fold cross validation is 91.8%, which is higher than other methods.