Table 1.

Summary of the experiments described in the manuscript together with their global settings

Tag	Cohort	Model integration a	Feature sets	Data sources
Classical b	Both	No	-	All c
Topological b	Both	Yes	All	All c
Integrated b	Both	Yes	All	All c
Centrality	Both	No	Centralities (all)	All c
Single centrality	Both	No	Centralities (one)	All c
node2vec	Both	No	node2vec	All c
Diffusion	Both	No	Diffusion	All c
Modularity	Both	No	Modularities	All c
Transcriptomic (microarray)	Both	No	All	Transcriptomic (microarray)
Transcriptomic (RNA-seq)	Both	No	All	Transcriptomic (RNA-seq)
Transcriptomic (both)	Small d	No	All	Transcriptomic (both)
Genomic (aCGH)	Small	No	All	Genomic
Fused	Both	Yes	All	All c

For the parameters that are not mentioned (e.g., dimension reduction strategy, network inference method, classification algorithm), the experiments are repeated for all possible values. ^aIntegration with weighted voting scheme. ^bAn equivalent tag for these models on the small cohort is All three sources. ^cThis means two on the large cohort and three on the small cohort. ^dOn the large cohort, it is equivalent to the topological model