Fig. 2
From: Predicting RNA sequence-structure likelihood via structure-aware deep learning
Example of 3D RNA matrix. The combination of four stacked subfigures on the top is the input of NU-ResNet. The flat form of four stacked subfigures are shown at the bottom. The 4 subfigures at the bottom represent the encoded \(G_{ij1}\), \(G_{ij2}\), \(G_{ij3}\), and \(G_{ij4}\), respectively. Fundamentally, each subfigure is a 2D matrix, which refers to a channel in the 3D RNA matrix. In this 3D RNA matrix example, the white box and black box correspond to 1 and 0 values of the matrix, respectively