Description

SVM model developed using single amino acid composition had an accuracy of 87.4% with 86.95% sensitivity and 87.4% specificity. The corresponding area under the curve was 0.95. Table 3.1 and Table 3.2 represents the number of features selected and the respective performance evaluation metrics obtained at various g-gaps for dipeptide and tripeptide respectively. The g-gap model with best accuracy is highlighted in Table 3.1 and Table 3.2. Using g-gap dipeptide compositional features, the overall accuracy of the models was in the range of 82-90% with the maximum accuracy of 90.83% obtained at gap-2. Similarly, for the models developed using g-gap tripeptide compositional features.