The Application of Support Vector Machine and Behavior Knowledge Space in the Disulfide Connectivity Prediction Problem
In this paper, we apply support vector machine (SVM) and behavior knowledge space (BKS) to the disulfide connectivity prediction problem. The problem aims to establish the disulfide connectivity pattern of the target protein. It is an important problem since a disulfide bond, formed by two oxidized cysteines, plays an important role in the protein folding and structure stability. The disulfide connectivity prediction problem is difficult because the number of possible patterns grows rapidly with respect to the number of cysteines. We discover some rules to discriminate the patterns with high accuracy in various methods. Then, the pattern-wise and pair-wise BKS methods to fuse multiple classifiers constructed by the SVM methods are proposed. Finally, the CSP (cysteine separation profile) method is also applied to form our hybrid method. We perform some simulation experiments with the 4-fold cross-validation on SP39 dataset. The prediction accuracy of our method is increased to 69.1 %, which is better than the best previous result 65.9 %.
KeywordsDisulfide bond Cysteine Connectivity pattern Support vector machine Behavior knowledge space
This research work was partially supported by the National Science Council of Taiwan under contract NSC 100-2221-E-242-003.
- 5.Baldi, P., Cheng, J., Vullo, A.: Large-scale prediction of disulphide bond connectivity. In: Saul, L., Weiss, Y., Bottou, L. (eds.) Advances in Neural Information Processing Systems, vol. 17, pp. 97–104. MIT Press, Cambridge (2005)Google Scholar
- 13.Frasconi, P., Passerini, A., Vullo, A.: A two-stage svm architecture for predicting the disulfide bonding state of cysteines. In: Proceedings of the IEEE Workshop on Neural Networks for Signal Processing, pp. 25–34 (2002)Google Scholar
- 19.Wang, C.-J., Yang, C.-B., Hor, C.-Y., Tseng, K.-T.: Disulfide bond prediction with hybrid models. In: Proceedings of the 2012 International Conference on Computing and Security (ICCS 2012), Ulaanbaatar, Mongolia, July 2012Google Scholar
- 20.Vapnik, V.N.: The Nature of Statistical Learning Theory. Springer, Heidelberg (1999)Google Scholar
- 21.Chang, C.-C., Lin, C.-J.: LIBSVM: A library for support vector machines (2001). http://www.csie.ntu.edu.tw/cjlin/libsvm
- 23.Chung, W.-C., Yang, C.-B., Hor, C.-Y.: An effective tuning method for cysteine state classification. In: Proceedings of National Computer Symposium, Workshop on Algorithms and Bioinformatics, Taipei, Taiwan, 27–28 November 2009Google Scholar
- 24.Chen, G., Deng, H., Gui, Y., Pan, Y., Wang, X.: Cysteine separations profiles on protein secondary structure infer disulfide connectivity. In: 2006 IEEE International Conference on Granular Computing, pp. 663–665, May 2006Google Scholar
- 29.Chen, Y.-C.: Prediction of Disulfide Connectivity from Protein Sequences. Ph.D. dissertation, National Chiao Tung University, Hsinchu, Taiwan (2007)Google Scholar