Fig. 5From: A systematic review of biologically-informed deep learning models for cancer: fundamental trends for encoding and interpreting oncology dataNetwork of relations between key components of the bio-centric interpretability. Network representing the relations between domain knowledge (red nodes), DK databases (orange), DK integration type according to the proposed taxonomy of bio-centric interpretability(purple nodes), DL models (blue nodes) and explainability methods (green nodes). Node size is proportional to the no. occurrences of the entity, edge width is proportional to no. pairs observed in the reviewed papers. We observe strong connections between: ARCH-pathways-sparse DNN; VAE-latent space exploration; PRE-PPI network-GCN; KEGG-pathwaysBack to article page