Skip to main content
. 2025 Aug 22;15(3):20250014. doi: 10.1098/rsfs.2025.0014

Figure 2.

Graph autoencoder for predicting additional connections in GRNs.

Graph autoencoder for predicting additional connections in GRNs. ( A) GAE network architecture. The encoder consists of a two-layer graphSAGE network. The decoder consists of an inner-product decoder with learnable weights. ( B) Procedure for improving the GRN. The GAE is trained for each edge for which a prediction is desired, and the results are subsequently combined to obtain an improved graph structure.