The collaboration between Nankai University, Shandong University, and Beijing Institute of Technology has led to the development of trRosettaRNA, an automated RNA 3D structure prediction method based on deep learning. This method employs a Transformer network to predict 1D and 2D geometric shapes, followed by energy minimization for 3D structure folding. In blind tests, trRosettaRNA has demonstrated strong competitiveness, particularly in performance evaluations compared to other deep learning methods. However, challenges remain in RNA 3D structure prediction, especially when dealing with large RNA structures and novel RNA types. Researchers hope that advancements in deep learning will enhance the accuracy and efficiency of RNA structure prediction.