Protein structure prediction is essential for comprehending protein function and developing therapeutic interventions. This paper focuses on review of various papers related to the protein structure prediction by using Deep Learning (DL) techniques and also performs a comparison of DL models employed in protein structure prediction, utilizing a dataset sourced from PDB. The paper covers DL principles, ensemble strategies and performs the comparative study of various DL techniques on the basis of various parameters such as accuracy, precision, recall, and F1 score. The findings provide nuanced insights into the strengths and limitations of each technique, contributing to the refinement of computational methods for the prediction of protein structure. This research aims to advance the field, offering a valuable foundation for the development of more accurate and scalable predictive models.