Skip to main content
. Author manuscript; available in PMC: 2024 Mar 14.
Published in final edited form as: Proteomics. 2023 Jun 27;23(17):e2200323. doi: 10.1002/pmic.202200323

Table 4.

list of the 20 raw scores with highest impact on the RF classification. The listed scores are the top 20 scores contributing to the RF classification of the benchmark dataset (see Figure 6).

Feature (raw score) Participant References
‘F_dG-cross/dSASAx100’ Furman RosettaCommon*, [71]
‘F_Voronota_iface_expanded_area’ Venclovas [21,85]
‘F_Voronota_diff_nonsas_area’
‘F_Voronota_diff_glob_light_score’
‘F_bsa’ Oliva NACCESS*, [86]
‘F_Voronota_diff_sas_area’ Venclovas [21,85]
‘F_BSA_Polar naccess’ Oliva NACCESS*, [86]
‘F_NIPacking’ Fernandez-Recio [58,87]
‘F_BSA_Polar_Freesasa’ Oliva NACCESS*, [86]
‘F_Voronota_diff_glob_dark_score’ Venclovas [21,85]
‘F_dG-separated/dSASAx100’ Furman
‘F_Voronota_iface_shelled_energy_norm’ Venclovas [21,85]
‘F_Voronota_iface_expanded_shelled_energy_norm’
‘F_Voronota_diff_glob_energy_norm’
‘F_dSASA-polar’ Furman RosettaCommon*, [71]
‘F_Voronota_iface_area’ Venclovas [21,85]
‘F_NSC’ Fernandez-Recio [58]
‘F_Voronota_iface-Atoms_light_score’ Venclovas [21,85]
‘F_CG_BETA’ Fernandez-Recio [58]
-F_Voronota_iface_enery_norm’ Venclovas [21,85]