Skip to main content

Table 4 Details about the architecture of Deep GONet

From: Deep GONet: self-explainable deep neural network based on Gene Ontology for phenotype prediction from gene expression data

Layer Input 1 2 3 4 5 6 Output Total
Level GO-BP 7 6 5 4 3 2 6
#neurons 54675 1574 1386 951 515 255 90 1 4772
#connections 86M 2.2M 1.3M 490K 131K 23K 90 90.1M
#GO connections 43504 1709 1585 1010 491 175 48K