From: GNE: a deep learning framework for gene network inference by aggregating biological information
Symbol | Definitions |
---|---|
M | Total number of genes in gene network |
E | Number of expression values for each gene |
N i | Set of the neighbor genes of gene vi |
\(\textbf {v}_{i}^{(s)}\) | Topological representation of gene vi |
\(\textbf {v}_{i}^{(a)}\) | Attribute representation of gene vi |
\(\widetilde {\textbf {v}}_{i}\) | Neighborhood representation of gene vi |
v i | Concatenated representation of topological properties and expression data |
k | Number of hidden layers to transform concatenated representation into embedding space |
h ( k) | Output of kth hidden layer |
W k | Weight matrix for kth hidden layer |
W id | Weight matrix for topological structure embedding |
W att | Weight matrix for attribute embedding |
W out | Weight matrix for output layer |