This review discusses the progress and applications of deep learning methods in protein structure prediction and design. It highlights the development of AlphaFold2 and other deep learning models that have significantly advanced the field. These models can predict the structures of individual proteins, protein complexes, and different conformations, with high accuracy. The review also explores the use of these models in evolutionary studies, protein design, and integrative structural modeling. Additionally, it discusses the potential extensions of these methods to other types of molecules and the integration of experimental data. The authors emphasize the importance of combining computational and experimental approaches to advance structural biology and provide a perspective on future directions in this field.This review discusses the progress and applications of deep learning methods in protein structure prediction and design. It highlights the development of AlphaFold2 and other deep learning models that have significantly advanced the field. These models can predict the structures of individual proteins, protein complexes, and different conformations, with high accuracy. The review also explores the use of these models in evolutionary studies, protein design, and integrative structural modeling. Additionally, it discusses the potential extensions of these methods to other types of molecules and the integration of experimental data. The authors emphasize the importance of combining computational and experimental approaches to advance structural biology and provide a perspective on future directions in this field.