Current research relies significantly on DNA sequencing. It enables the advancement of a wide range of disciplines, including genomics, metagenomics, and phylogeny. DNA sequencing involves extracting and interpreting DNA strands. In recent years, algorithms based on machine learning have been effectively employed to identify and classify viruses, which is crucial for preventing outbreaks such as COVID-19. Feature extraction aids in identifying the impact of viruses and designing drugs. The research presented here compares DNA sequencing employing Machine Learning algorithms, Deep Learning methods, and Hybrid algorithms. The objective of our suggested approach is to create an improved framework of prediction for DNA studies and achieve the most accurate findings possible. Machine learning, deep learning, and hybrid approaches are the most widely utilized and well-known techniques in this field. Additionally, deep learning provides higher predictive accuracy, and has demonstrated to perform better in several medical fields. The suggested frameworks are Decision Tree, Random Forest, Naive Bayes, Transform Learning, Convolutional Neural Network (CNN), Convolutional Neural Networks-Long Short-Term Memory (CNN-LSTM), as well as Bidirectional Long Short-Term Memory (Bidirectional LSTM). The CNN approach provided 98.4% accuracy with label encoding during training and 96.98% during testing, whereas the Bidirectional LSTM algorithm achieved 98.13% accuracy using K-mer encoding during testing and 97.7% during training. CNN-LSTM with K-mer encoding achieved 97.69% accuracy on collected data during testing and 96.23% during training.