Overview of traditional and modern deep learning methods for protein structure prediction (IMAGE)
Caption
This review provides an overview of traditional and modern methods for protein structure prediction and their characteristics and introduces the groundbreaking network features of the AlphaFold family model. We summarize the applications of different bioinformatics databases and various models based on deep learning architectures (CNN, RNN, Transformer, etc.) in this field, which may provide directions for future development of predictive models such as prediction of the structure of large protein complexes and integration of protein language models.
Credit
Not applicable
Usage Restrictions
Not aplicable
License
Original content