Many diseases (including cancer) are related to non-coding RNAs, but we don't know much about them because the structures used to understand their functional mechanisms are difficult to resolve by physical methods such as X-ray crystal diffraction, nuclear magnetic resonance, and cryo-electron microscopy. Without the understanding of the mechanism, there will be no means to completely solve the disease, so it is imperative to predict the structure of RNA through calculations and greatly improve its accuracy. For a long time, the prediction of RNA structure has focused on secondary structure (base pairing) and used it for the constraint of tertiary structure prediction. To optimize the energy function that is not very reliable, Yaoqi Zhou and the Paliwal research group of Griffith University developed a tool that uses deep learning to predict the dihedral angle of the main chain will change this state. This pioneering result was published in J. Chem . Info. Modeling (2021). For details, please click: News link