From: Ranking near-native candidate protein structures via random forest classification
Data set
Number of proteins
Average length
I-TASSER Decoy Set-I
43
80
QUARK Decoy Set
145
107
CASP10 dataset
54
212
CASP11 dataset
39
203