Assignment of protein sequence to functional family using neural network & Dempster-Shafer theory

Abstract

Protein classification prediction is an important problem in molecular biology, and one that has attracted a lot of attention. This paper describes an approach to data-driven discovery of sequence motif-based models using neural network classifier based on Dempster-Shafer Theory for assigning protein sequences to functional families. A training set of sequences with unknown functional family is used to capture regularities that are sufficient to assign the sequences to their respective families. A new adaptive pattern classifier based on neural network and Dempster-Shafer theory of evidence developed by Thierry Denoux, 2001, [2] is presented. This method uses reference patterns as items of evidence regarding the class membership of each input pattern under consideration. This evidence is represented by basic belief assignments (BBA's) and pooled using the Dempster's rule of combination. This procedure can be implemented in a multilayer neural network with specific architecture consisting of one input layer, two hidden layers and one output layer. The weight vector, the receptive field and the class membership of each prototype are determined by minimizing the mean squared differences between the classifier outputs and target values.