Skip to main content

Table 5 The result of 10 fold cross validation using heat diffusion algorithm with baseline prediction

From: Predicting links between tumor samples and genes using 2-Layered graph based diffusion approach

Method AUC-ROC AUC-PR
Common neighbor approach (CN) 0.72 0.0382
Jaccard similarity (JS) 0.76 0.0549
Preferential attachment (PA) 0.73 0.0428
Resource allocation (RA) 0.78 0.0612
Adamic adar index (AAI) 0.79 0.0645
Katz (STRING) α=0.15,β=0.0001 0.54 0.0161
Katz (BioGRID) α=0.15,β=0.0001 0.72 0.0311
Personalized pageRank algorithm (PPR) (STRING) 0.81 0.0523
Personalized PageRank algorithm (PPR) (BioGRID) 0.81 0.0543
Heat diffusion algorithm (HD) (STRING) 0.85 0.0823
Heat diffusion algorithm (HD) (BioGRID) 0.74 0.0321
Random baseline (RB) 0.5 0.0125
Random network in layer 2 (RN) 0.5 0.0012
\