Skip to main content
. 2024 Nov 26;26(1):bbae611. doi: 10.1093/bib/bbae611

Figure 1.

Figure 1

Framework of LIMO-GCN. (A) The input of LIMO-GCN. It contains an adjacency matrix generated from AD-specific FGN and a feature matrix X, which is comprised of PPI from STRING database and molecular signatures from MSigDB. (B) The linear model of LIMO-GCN. (C) The GCN model of LIMO-GCN. (D) The output of vector y is taken as predicted score. Gene with a darker color indicates a higher association score with AD.