. 2009 May 20;9:35. doi: 10.1186/1472-6807-9-35

Table 6.

Comparison of consensus and non-consensus based methods in the estimation of the local model quality.

Scoring function	r	tau	ROC_avg	ROC_all	low_10%	top_10%
QMEANclust_local	0.83	0.53	0.88	0.93	2.2	29.5
selfQMEAN_local	0.49	0.35	0.84	0.90	1.3	5.8
QMEAN_local	0.43	0.32	0.80	0.83	0.8	4.3
ProQres	0.28	0.26	0.74	0.77	0.9	5.8

r = average Pearson's correlation coefficient; tau = Kendalls's tau on a per model basis; ROC = area under ROC curve averaged over all 98 targets (avg) or using all residues pooled together (all); low/top10% = average Cα distance of the 10% lowest/highest scoring residues per target.