Features Extraction Using Free Score of Words for Classifying Conotoxin Superfamily

Interest in Conotoxin has been rapidly growing over the past number of years due to its potential for effective use in the design of drugs to treat a myriad of conditions including, neuromuscular disorders, chronic pain and schizophrenia. As a result it is necessary to develop powerful and efficient techniques which can accurately classify conotoxin super families. In this paper, we propose a novel technique which makes use of support vector machines for classification. The method which considers suboptimal alignments of words with restricted length and computes local alignment partition functions to produce free scores for the alignments plays the key role in the feature extraction step of support vector machine classification. In the classification of conotoxin proteins, the proposed approach, SVM-Freescore, demonstrates its potential use by yielding an improved sensitivity and specificity of approximately 5.864% and 3.76%, respectively.