Prediction of disulfide bonding pattern based on support vector machine with parameters tuned by multiple trajectory search

Authors:
Hsuan-Hung Lin;Lin-Yu Tseng
Affiliations:
Department of Applied Mathematics, National Chung Hsing University, Taichung, Taiwan, ROC and Department of Management Information Systems, Central Taiwan University of Science and Technology, Tai ...;Institute of Networking and Multimedia, National Chung Hsing University, Taichung, Taiwan, ROC and Department of Computer Science and Engineering, National Chung Hsing University, Taichung, Taiwan ...
Venue:
WSEAS Transactions on Computers
Year:
2009

Citing 17
Cited 0

An introduction to support Vector Machines: and other kernel-based learning methods

An introduction to support Vector Machines: and other kernel-based learning methods
Training Invariant Support Vector Machines

Machine Learning
Implementation of algorithms for maximum matching on nonbipartite graphs.

Implementation of algorithms for maximum matching on nonbipartite graphs.
Disulfide connectivity prediction using recursive neural networks and evolutionary information

Bioinformatics
Cysteine separations profiles on protein sequences infer disulfide connectivity

Bioinformatics
Disulfide connectivity prediction using secondary structure information and diresidue frequencies

Bioinformatics
A hybrid machine-learning approach for segmentation of protein localization data

Bioinformatics
Improving disulfide connectivity prediction with sequential distance between oxidized cysteines

Bioinformatics
A support vector machine-based method for predicting the propensity of a protein to be soluble or to form inclusion body on overexpression in Escherichia coli

Bioinformatics
Predicting disulfide connectivity from protein sequence using multiple sequence feature vectors and secondary structure

Bioinformatics
Predicting disulfide bond connectivity in proteins by correlated mutations analysis

Bioinformatics
RISP: A web-based server for prediction of RNA-binding sites in proteins

Computer Methods and Programs in Biomedicine
Comparison studies on classification for remote sensing image based on data mining method

WSEAS Transactions on Computers
Electrical energy consumption forecasting based on cointegration and a support vector machine in China

WSEAS Transactions on Mathematics
Mammographic masses characterization based on localized texture and dataset fractal analysis using linear, neural and support vector machine classifiers

Artificial Intelligence in Medicine
Multiple trajectory search for unconstrained/constrained multi-objective optimization

CEC'09 Proceedings of the Eleventh conference on Congress on Evolutionary Computation
A comparison of methods for multiclass support vector machines

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

The prediction of the location of disulfide bridges helps solving the protein folding problem. Most of previous works on disulfide connectivity pattern prediction use the prior knowledge of the bonding state of cysteines. In this study an effective method is proposed to predict disulfide connectivity pattern without the prior knowledge of cysteins' bonding state. To the best of our knowledge, without the prior knowledge of the bonding state of cysteines, the best accuracy rate reported in the literature for the prediction of the overall disulfide connectivity pattern (Qp) and that of disulfide bridge prediction (Qc) are 48% and 51% respectively for the dataset SPX. In this study, the cystein position difference, the cystein index difference, the predicted secondary structure of protein and the PSSM score are used as features. The support vector machine (SVM) is trained to compute the connectivity probabilities of cysteine pairs. An evolutionary algorithm called the multiple trajectory search (MTS) is integrated with the SVM training to tune the parameters for the SVM and the window sizes for the predicted secondary structure and the PSSM. The maximum weight perfect matching algorithm is then used to find the disulfide connectivity pattern. Testing our method on the same dataset SPX, the accuracy rates are 54.5% and 60% for disulfide connectivity pattern prediction and disulfide bridge prediction when the bonding state of cysteines is not known in advance.