ABSTRACT
Dengue virus poses a significant global health challenge, necessitating a deeper understanding and innovative therapeutic interventions. The present study proposes a novel computational approach employing Machine Learning (ML) to enhance the efficiency of conventional vaccine discovery processes. Approximately 1.3 billion individuals residing in Southeast Asia (SEA) are at risk of contracting dengue virus infection. Notably, five SEA nations (Indonesia, Burma, Sri Lanka, Thailand, and India) are among the top 30 most highly endemic countries worldwide. Over time, the number of dengue cases has increased significantly, despite efforts to control the disease. Between 2015 and 2019, the SEA Region witnessed a 46% increase in dengue cases, while dengue-related deaths decreased by 2%. Several factors contribute to the spread and distribution of dengue mosquito vectors and viruses in the SEA Region. The current high burden of dengue cases in the SEA Region is exacerbated by the lack of effective treatments and sustainable vector control measures. The objective is to identify and develop a multi-epitope vaccine as a weapon against antigenic proteins of DENV3. The traditional vaccine development process is very time-consuming. By employing an in-silico (computational) vaccine discovery approach integrated with machine learning (ML) algorithms like artificial neural networks (ANN), random forest (RF), and decision trees, the time and need for resources is drastically reduced. These ML models can be trained on experimental data to aid in the design of multi-epitope vaccine to combat the DENV3 dengue virus serotype.
