Self-Organizing Maps
Learning activity patterns using fuzzy self-organizing neural network
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Scoring hidden Markov models to discriminate β-barrel membrane proteins
Computational Biology and Chemistry
A HMM-based method to predict the transmembrane regions of β-barrel membrane proteins
Computational Biology and Chemistry
| Hi-index | 0.01 |
A novel method based on fuzzy Self-Organizing Map to detect the transmembrane segments, called TSFSOM, is presented in the paper. The multivariate ”time” series of transmembrane proteins are classified by fuzzy Self-Organizing Map into five classes. Through the analysis of resulting trajectories on the map, frequent patterns of transmembrane segments are detected and even some kind of ”new” knowledge about membrane insertion mechanism is obtained. The discovered patterns and the knowledge are then used to predict transmembrane segments for query sequence. The prediction results not only show that the method is powerful, but also prove that the patterns and the knowledge about the interaction between the patterns are effective and acceptable.