How does WGCNA decide which genes are connected?

Question

[jolespin]

I'm using the Weighted Gene Coexpression Network Analysis package (WGCNA) in R and I am trying to figure out how the adjacency matrix is defined.

I understand how the correlation matrix is created along with why the power is used but adjacency matrices should be either 1 or 0.

Why are there still continuous values? How does it choose a cutoff later?

I can understand that it's maybe a weighted adjacency matrix but there isn't an option to make it binary and that is needed for the topological overlap matrix calculation (TOM)

Answer 1

[sina.nassiri]

An adjacency matrix is a square, symmetric matrix with entries between 0 and 1 (don't have to be 1s and 0s), and the TOMsimilarity() and TOMdist() functions in WGCNA both handle non-binary adjacency matrices.