Skip to main content

Advertisement

Fig. 5 | BMC Bioinformatics

Fig. 5

From: PASNet: pathway-associated sparse deep neural network for prognosis prediction from high-throughput data

Fig. 5

Hierarchical representation of pathways in PASNet. (a) PASNet is partially visualized showing the five pathways. Distinct neural network activations between LTS (b) and non-LTS (c) are shown via PASNet. The nodes of the neural network of (b) and (c) correspond to (a). For instance, the nodes in the pathway layer of (b) and (c) represent signaling by GPCR, innate immune system, aquaporin-mediated transport, signaling by BMP, and Cytokine signaling in immune system. The pathways of signaling by GPCR and innate immune system are inactive with LTS patients, whereas the both pathways are active with non-LTS patients

Back to article page