Advanced Detection and Classification of COVID Variants Based on Omics Approach and Machine Learning Models

The COVID-19 pandemic, driven by SARS-CoV-2, highlighted the critical need for swift and precise classification of viral variants to enhance public health measures. This research introduces a machine learning method for identifying coronavirus variants such as Alpha, Delta, and Omicron, achieving accuracy rates of up to 99%. Through feature selection methods like Mutual Information and ANOVA, along with advanced classifiers such as Gradient Boosting, Random Forest, and CatBoost, we pinpointed crucial amino acids that differentiate these variants. Our results demonstrate that machine learning improves classification accuracy, contributes to understanding the evolution of variants, and facilitates scalable viral genomics. This approach holds significant potential for advancing vaccine development and allowing for prompt, data-informed interventions.