Table 2.

Statistics of the datasets, where the Density is defined as $2\times \text{no}. \text{edges}/\text{no}. \text{node}{\mathrm{s}}^2$

Task type	Dataset	No. nodes	No. edges	Density	No. node labels
Link prediction	CTD DDA	12 765	92 813	0.11%	—
	NDFRT DDA	13 545	56 515	0.06%	—
	DrugBank DDI	2191	242 027	10.08%	—
	STRING PPI	15 131	359 776	0.31%	—
Node classification	Clin Term COOC	48 651	1 659 249	0.14%	31
	node2vec PPI	3890	76 584	1.01%	50
	MashUp PPI	16 143	300 181	0.23%	28