Skip to main content

Table 3 The rankings of the 19 selected features

From: Computationally identifying hot spots in protein-DNA binding interfaces using an ensemble approach

Rank Feature name Type Rank Feature name Type
1 PSSM(R) Sequence 11 Lse score Sequence
2 H-Bond in HBPLUS Structure 12 phi in SPOT-1D Sequence
3 ALPHA in Xssp Structure 13 COMBINED2 score in ENDES Structure
4 Current_flow_closeness_centrality Network 14 ACC in Xssp Structure
5 Q3_prob_3 in NetSurfp2 Sequence 15 RSA in NetSurfp2 Sequence
6 HSEa-u in SPOT-1D Exposure 16 P(8-G) in SPOT-1D Sequence
7 COMBINED1 score in ENDES Structure 17 CN in hsexpo Exposure
8 SIDESCORE score in ENDES Structure 18 Conservation score Sequence
9 Q8_prob_1 in NetSurfp2 Sequence 19 Blosum(E) Sequence
10 P(8-I) in SPOT-1D Sequence    
  1. These features fall intro four types, i.e., network features, exposure features, structure features and sequence features