Table 1.

Summary of binary classification tasks and datasets

Type	Dataset
		Edge classification
		$\left|V\right|$	$\left|E\right|$	$n_{+}$	$n_{-}$
PPI	PC	4761	22 988	10 517	12 471
DTI	ID	544 drugs	10 436	4284	6152
		2261 targets
	AI	378 drugs	1039	249	790
		267 targets
		Hyperedge classification
		$\left|V\right|$	$\left|E\right|$	$n_{+}$	$n_{-}$
PPI	BC	3436	2357	145	161
	MP	3436	2357	175	200
		Link prediction
		$\left|V\right|$	$\left|E\right|$	$\left|V^{\text{lcc}}\right|$	|Elcc|a
PPI	EC	393	391	100	153
	CE	3026	5163	2779	5014
	AT	5391	12 825	5063	12 631
	SP	853	1197	685	1092
	RN	526	532	301	388
	MM	2065	2833	1590	2522
DTI	EZ	445 drugs	2926	809	2556
		664 targets
	IC	210 drugs	1476	409	1473
		204 targets
	GR	223 drugs	635	240	570
		95 targets
	NR	54 drugs	90	42	50
		26 targets

Note: For each learning problem, we show the number of vertices (V) and edges (E) in the full hypergraph, as well as the largest connected component ($V^{\text{lcc}},\hspace{0.17em}{E}^{\text{lcc}}$). We also show the number of positive ($n_{+}$), negative ($n_{-}$) or unlabeled (n_u) data points.

^aThe size of $n_{+}$ and n_u is given by $\left|E^{\text{lcc}}\right|$.