Skip to main content


Fig. 6 | BMC Bioinformatics

Fig. 6

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

Fig. 6

Architecture of PASNet. The structure of PASNet is constructed by a gene layer (an input layer), a pathway layer that represents the biological pathways linked with input genes, a hidden layer that represents hierarchical relationships among biological pathways, and an output layer that corresponds with clinical outcomes, e.g. a binary class that has long-term survival and short-term survival, stages of cancer

Back to article page