ProtNN: fast and accurate protein 3D-structure classification in structural and topological space

Table 1 Characteristics of the experimental datasets

Dataset	SCOP ID	Family name	Pos.	Neg.	Avg. ∣V∣	Avg. ∣E∣	Max. ∣V∣	Max. ∣E∣
DS1	48623	Vertebrate phospholipase A2	29	29	160	628	451	1812
DS2	52592	G-proteins	33	33	246	971	897	3544
DS3	48942	C1-set domains	38	38	238	928	768	2962
DS4	56437	C-type lectin domains	38	38	185	719	775	3016
DS5	56251	Proteasome subunits	35	35	231	929	897	3544
DS6	88854	Protein kinases, catalyc subunits	41	41	275	1077	775	3016

SCOP ID, Family name, Pos., Neg., Avg. ∣V∣, Avg. ∣E∣, Max. ∣V∣ and Max. ∣E∣ correspond respectively to the identifier of the positive protein family in SCOP, its name, the number of positive examples, the number of negative examples, the average number of nodes, the average number of edges, the maximal number of nodes and the maximal number of edges in each dataset

ISSN: 1756-0381