Type-2 fuzzy support vector machine model for conformational epitope prediction

Abstract

Identifying and predicting epitopes by experimental approaches is a time-consuming and expensive procedure. As a result, computational methods have been explored as a faster and more cost-effective alternative. Nevertheless, existing computational methods encounter difficulties in achieving accuracy due to the presence of data ambiguity. The type-1 fuzzy set may effectively manage the uncertainty present in data. Nevertheless, it lacks the ability to effectively manage uncertainty in the data’s relationship. This research proposes a model that combines a type-2 fuzzy set with a support vector machine to handle ambiguity in data relationships to enhance the accuracy of predicting conformational epitopes. The results obtained from the proposed method demonstrated a substantial enhancement in accuracy when compared to earlier methods in the prediction of conformational epitopes. © The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2024.