Support vector machine method on predicting resistance gene against Xanthomonas oryzae pv. oryzae in rice

Motivation fication of disease-resistant genes in the rice is a tough work in various experimental studies. Xanthomonas oryzae pv. oryzae (Xoo) which causes bacterial blight are considered to be the most devastating diseases in most rice-growing regions. However, currently there is no existing method for the prediction of disease-resistant genes from sequence data. Accurate prediction of Xoo from protein sequences is illuminating for gene finding projects. s pose a novel machine-learning approach based on the method of support vector machine (SVM) and chaos game representation (CGR), to assess the chance of a protein in rice to be Xoo resistant. We choose 13 already cloned genes for positive data and 48 selective gene in rice for negative data, the average accuracy achieves 100% in resubstitution test, 95.08% in jackknife test, and the Matthews correlation coefficient achieves 0.8509. The successful application of SVM + CGR approach in this study suggests that it should be more useful in quantifying the protein sequence–structure relationship and predicting the structural property profiles from protein sequences.