Mr Toheeb Jumah
University Mohammed VI Polytechnic (UM6P), Morocco
IMMET: A Machine Learning Model for Predicting Epitopes of New Strains of Mycobacterium TB and Aiding Immunogen Design Using Known and Current Mycobacterium tuberculosis Epitopes
Poster Abstract
The development of new, effective vaccines for Mycobacterium tuberculosis is essential for reducing the global burden of this disease. Novel tuberculosis vaccines focus on boosting the adaptive immune system by targeting specific antigens of Mycobacterium tuberculosis, which contain immunogenic epitopes capable of activating immune responses. These vaccines specifically aim to induce the activation of long-lived memory cells, which play a crucial role in the effectiveness of any vaccine.
Machine learning models that can predict new epitopes are crucial for vaccine development as they streamline the epitope discovery process, reducing the need for traditional, resource-intensive methods. To train an effective model, we used several supervised learning classification methods, including support vector machines (SVM) and random forests. We found that different models exhibit varying degrees of accuracy.
We used Mycobacterium tuberculosis epitopes, reported as being recognized during infection, as the datasets for the models to process. The findings of our study demonstrate the ability of machine learning algorithms to accurately predict epitopes of Mycobacterium tuberculosis. The support vector machine model showed remarkable accuracy in this categorization challenge and offers valuable insight into the mechanisms that aid the immune system in recognizing epitopes.
Utilizing machine learning techniques, we propose that we specifically improve prediction of Mycobacterium tuberculosis epitopes, and that this will prove to be a valuable tool in accelerating vaccine development and potentially tailoring therapeutic treatments. Further research aims at enhancing these models and advancing their capabilities in predicting the immunogenic nature of Mycobacterium tuberculosis epitopes for vaccine development.
