Design and Development of Computationally Intelligent Framework for Protein Allergen Classification

Abstract

The prevalence of allergic reactions is a serious public health problem that may cause diseases like high fever, rhinitis, asthma, dermatitis etc., and affects a sizeable fraction of the world’s population. Protein allergens are a key contributor to the development of allergic reactions; hence, locating and describing these allergens is essential to the research and development of efficient diagnostic and therapeutic methods. Currently, the methods that can be used to cure allergies are not completely understood, and the only strategy that is known for preventing allergies is to abstain from substances that contain allergens. The traditional techniques of allergen prediction take a lot of time and frequently depend on methods of experimentation that require a lot of manual resources. Also, the inaccurate detection of allergens may result in excessive dietary restrictions, which can then lead to nutritional issues. Hence, it is clear that an efficient mechanism that can amalgamate multiple parameters for accurate allergen detection is necessary for allergy control. The computational methods combined with bioinformatics have the ability to analyze multiple characteristics of sequences and structures of protein allergens for detecting potentially allergenic areas with better detection accuracy. The work presented in this thesis describes a thorough analysis of various state-of-the-art literature mechanisms and related approaches, resulting in the creation of various proposed mechanisms for the detection, prevention and control of allergies. The mechanisms proposed in the thesis used various computationally intelligent approaches like machine learning, deep learning, ensemble learning and stochastic modelling. A deep learning based ensemble approach for protein allergen classification is proposed in the thesis and has an accuracy of 89.16%, which is higher than other related work in the field. A plant and animal food allergen classification using extra tree based ensemble model is also presented, and the proposed model achieved an accuracy of 94.23% with 10 fold cross-validation. A numerical simulations of a six-compartment model was also conducted in order to predict the likelihood of developing allergies. The numerical presentation of the system’s solution was achieved through the utilisation of the stochastic computational artificial neural network (ANN) and the Levenberg-Marquardt backpropagation (LMBP). The validation of the model was achieved by comparing the results with a reference model. The overall combination of various proposed mechanisms and techniques in the thesis presents a computationally intelligent framework for the efficient detection of protein allergens.