AUTHOR=Tian Hao , Jiang Xi , Trozzi Francesco , Xiao Sian , Larson Eric C. , Tao Peng TITLE=Explore Protein Conformational Space With Variational Autoencoder JOURNAL=Frontiers in Molecular Biosciences VOLUME=Volume 8 - 2021 YEAR=2021 URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2021.781635 DOI=10.3389/fmolb.2021.781635 ISSN=2296-889X ABSTRACT=Molecular dynamic simulations have been extensively used in the studies of protein dynamical and functional properties. However, extensive and sufficient sampling in the protein conformational space requires large amount of computational powers and time. In this study, we demonstrated that variational autoencoders, a type of deep learning model, can be employed to explore protein conformational space through combination of molecular dynamics simulations. Variational autoencoders are shown to be superior to autoencoders through a benchmark study, with low deviation between the training and decoded conformations. Moreover, we showed that the learned latent space in the trained variational autoencoder can be used to generate unsampled protein conformations. Additional simulations starting from these generated conformations greatly enhanced the sampling efficiency and explored hidden spaces in the protein conformational landscape.