Table 2.

PredUs prediction performance averages when using structure neighbors from the same and different SCOP groupings on the docking benchmark dataset

Prediction methods	Cases	Precision average (%)	Recall average (%)
PredUs(server)	185	50.3	57.5
PredUs(original)	185	43.6	45.7
Family	141	50.8	33.8
Superfamily	147	45.9	36.2
Fold	153	41.8	38.7

Quantities in each column are defined in the description of the PredUs benchmarks in the main text.