Protein structure prediction has become extremely accurate, and its results are now comparable with those of experimental methods for a large number of proteins. However, there remain some technical hurdles to clear before the current structure prediction tools can be directly applied to a wide range of biomedical problems. New perspectives on future developments in the area of structure prediction and its biomedical applications are presented.
INTRODUCTION: A NEW ERA OF ACCURATE STRUCTURE PREDICTION BY ARTIFICIAL INTELLIGENCE
AlphaFold AND RoseTTAFold: LINKING COEVOLUTIONARY INFORMATION IN PROTEIN SEQUENCES TO PROTEIN STRUCTURE USING DEEP NEURAL NETWORKS
ACCURATE PROTEIN STRUCTURE PREDICTION: LIMITATIONS AND OPPORTUNITIES
BARRIERS TO ACCURATE PREDICTION FOR FUTURE BIOMEDICAL APPLICATIONS
SPECULATIONS FOR FUTURE DEVELOPMENT IN PROTEIN MODELING BY ARTIFICIAL INTELLIGENCE
ACKNOWLEDGEMENTS
CONFLICT OF INTEREST
REFERENCES