Table 1. List of methods included in this benchmark.

Method Identifier	Method Name	Method Class	Implementation	Reference
pr	PageRank with a uniform prior	Baseline	igraph (Bioconductor [24, 25] package)	[26]
random	Random	Baseline	R	(see text)
randomraw	Random Raw	Baseline	R	(see text)
EGAD	Extending Guilt by Association’ by Degree	Baseline	EGAD (Bioconductor package)	[27]
ppr	Personalized PageRank	Diffusion	igraph (R package)	[28]
raw	Raw Diffusion	Diffusion	diffuStats (Bioconductor package)	[29]
gm	GeneMania-based weights	Diffusion	diffuStats (Bioconductor package)	[30]
mc	Monte Carlo normalised scores	Diffusion	diffuStats (Bioconductor package)	[31]
z	Z-scores	Diffusion	diffuStats (Bioconductor package)	[31]
knn	K nearest neighbours	Semi-supervised learning	RANKS (R package)	[32]
wsld	Weighted Sum with Linear Decay	Semi-supervised learning	RANKS (R package)	[32]
COSNet	COst Sensitive neural Network	Supervised learning	COSNet (R package)	[33]
bagsvm	Bagging SVM (based on ProDiGe1)	Supervised learning	kernlab (R package)	[34]
rf	Random Forest	Supervised learning	randomForest (R package) + Matlab (features)	[35]
svm	Support Vector Machine	Supervised learning	kernlab (R package) + Matlab (features)	[35]

Method identifiers are shortened method names used throughout the text. Other columns are self-explanatory.