Figure 7
From: Using random walks to identify cancer-associated modules in expression data
Comparison of top modules from Walktrap-GM , Matisse, jActiveModules . Performance in finding modules significantly enriched with known cancer genes,across breast cancer (BC) and hepatocellular carcinoma (HCC) and colorectal cancer data (CCA). Green lines show Walktrap-GM performance, blue jActiveModules, and orange Matisse. Walktrap-GM performs as well as or better than the other approaches across datasets. In the BC data, blue jActiveModules resulted in one very large and significant module of 981 nodes, but few significant modules overall. Matisse includes overlapping significant genes within its modules where Walktrap-GM does not and jActiveModules is configured not to inlcude overlap.