A framework for protein structure classification and identification of novel protein structures

Table 3 The comparison between SCOPmap and proCC using the predicted Superfamily SCOP labels

	Classified with correct domain boundary		Unclassified with correct domain boundary		Incorrect domain boundary	Overall precision	New class detection ratio	Estimated average execution time
	Correct CC	Incorrect CI	New classes UN	Existing classes UE	ID	$\frac{(C C + U N)}{2773}$	$\frac{UN}{307}$
SCOPmap	2069	65	190	212	237	81.5%	61.9%	2–3 hours per query
proCC	2025	75	246	275	152	81.9%	80.1%	9 minutes per query

This table shows the result of classifying 2773 single domain chains in SCOP 1.69. All numbers reported in column 2–6 are in terms of the number of chains (or domains due to the fact that we used single domain chains). Column 2–3 show the number of single domain chains which are correctly identified as single domain chains and are classified to known superfamilies. Column 4–5 show the number of single domain chains which are correctly identified as single domain chains and are labeled as unclassified. Column 6 shows the number of single chain domains which are incorrectly identified as multi-domain chains.