Skip to main content


Figure 5 | BioData Mining

Figure 5

From: The multiscale backbone of the human phenotype network based on biological pathways

Figure 5

How to choose α . The networks average clustering coefficient (CC), percentage of conserved vertices (% vertices), percentage of conserved edge weights (% weights) and, in (a) and (b) the value of α as functions of the percentage of remaining edges (% edges). (a) and (b) α varies in [0,1]. We chose α=0.25, which yields the best results, % edges =0.1, and use this value in (c) to identify the global weight cutoff =292. Note that curves are reported as functions of % edges, not α, to allow direct comparison between the DF in (a) and (b) and the GW filter in (c).

Back to article page