Name
Abstract | ||
|---|---|---|
Locating the disulfide bridges is an important task in protein structure prediction. The correct prediction of disulfide bridges can reduce the conformational search space and improve the results while predicting protein structures. This study predicts disulfide connectivity patterns with prior knowledge of binding states. A Support vector machine (SVM) is adopted herein to generate bond potentials. These bond potentials are then mapped to a weighted graph to find the perfect maximal weight match by the Edmonds' algorithm. Different features are considered for predicting bond potentials, including sliding window of sequence, predicted secondary structure, hydrophobicity, and sequential distance. A 4-fold cross-validation procedure on a data set containing 452 proteins is performed to validate the proposed approach. As a result, the overall accuracy of the proposed approach is 45%, which is comparable or even better than previous works. The inferences of features are also discussed. |
Year | Venue | Keywords |
|---|---|---|
2004 | METMBS '04: PROCEEDINGS OF THE INTERNATIONAL CONFERENCE ON MATHEMATICS AND ENGINEERING TECHNIQUES IN MEDICINE AND BIOLOGICAL SCIENCES | Support Vector Machine,disulfide-bond connectivity |
Field | DocType | Citations |
Pattern recognition,Disulfide bond,Computer science,Support vector machine,Artificial intelligence | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Chi-Hung Tsai | 1 | 116 | 8.70 |
| Huai-Kuang Tsai | 2 | 132 | 14.33 |
| Shih-chieh Chen | 3 | 0 | 0.34 |
| Cheng-yan Kao | 4 | 586 | 61.50 |