Deep learning for reconstructing protein structures from cryo-EM density maps: recent advances and future directions

Deep learning for reconstructing protein structures from cryo-EM density maps: recent advances and future directions Nabin Giri, Raj S. Roy, Jianlin Cheng Deep learning is a promising technique for efficient, automatic, and accurate reconstruction of protein structures from cryo-EM density maps Advanced convolutional neural networks and U-Nets have been successfully applied to reconstruct protein structures from high-resolution cryo-EM density maps Creating high-quality cryo-EM data sets for training and testing deep learning methods is important and there is a significant need of curating such data sets to facilitate the development of deep learning methods Better structure reconstruction can be obtained by combining AlphaFold predicted structure models and cryo-EM data and by integrating cryo-EM based structure determination techniques and protein structure prediction techniques.