Table 1 Properties of full-size and reduced graphs.

From: An integrated network representation of multiple cancer-specific data for graph-based machine learning

Property

Full-size graph

Reduced graphs

Number of nodes

19,144

1349 ± 80

Number of edges

685,198

12,613 ± 608

Average degree

71

19 ± 0.3

Density

0.004

0.014 ± 0.0009

Diameter

8

4.073 ± 0.26

Clustering coefficient

0.287

0.659 ± 0.006

Maximum betweenness centrality

0.021

0.596 ± 0.011

Average betweenness centrality

1.11 × 10−4

7.88 × 10−4 ± 4.49 × 10−6

  1. Statistics are calculated from the graph topology without considering node features. Values for reduced graphs are reported as the average ±standard deviation across the dataset.
Back to article page