Predicting subcellular localization of apoptosis protein based on ensemble learn-ing and Gene Ontology annotation database
WANG Xiao
LI Hui
ZHAI Yun-qing
Abstract：In order to deal with the problem that the prediction accuracy of subcellular localization of apoptosis proteins is not high,a method of predicting subcellular localization of apoptosis protein based on ensemble learning and Gene Ontology (GO)annotation database was proposed.This method utilized the GO features of apoptosis proteins and their homologous proteins combined with the two layer integration strategy to predict subcellular localization of apoptosis proteins.In the first layer,several sets of feature vectors were formulated by the different number of homologous proteins,then it selected the distance weighted K-nearest neighbor clas-sifier as individual classifier,trained sub-prediction models,and integrated these models by majority voting. In the second layer,the prediction model of the first layer was used as the sub-prediction model,and it inte-grated the different nearest neighbors’sub-prediction models by the majority voting.The results of Jackknife test showed that prediction accuracy of the method reaches 96.2% on the CL31 7 apoptosis proteins dataset, which was superior to other methods.In addition,this method could reduce the impact of the data imbalance.
Keywords：subcellular locationhomologous proteinsGene Ontology(GO) featuresensemble K nearest-neighbor algorithm
Publication Date：2016-07-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：7( 95-101 )