Table 2.
Year | Name | Detection | Principle | Scoring | Evaluation Datasets |
---|---|---|---|---|---|
2002 | CavBase [17] | LIGSITE [12] | Clique detection in graphs of pseudoatoms | Overlap of surface grid points, RMSD | Cofactor sites, kinases, serine proteases |
2002 | eF-site [109] | Ligand Databases |
Clique detection in graph of surface normal vectors and electrostatic potentials | Normalized and weighed contributions of vectors angles, potentials, distances | Phosphate sites, antibodies, PROSITE classes |
2003 | SuMo [110] | Ligand | Incremental match of triplets of pseudocenters | Count of matches, RMSD, composite of Euclidean and density distances | Protease catalytic sites, lectines |
2004 | SiteEngine [18] | Ligand | Match of triplets of points by hashing | Hierarchical scoring: count of matches, RMSD, overlap of patches, local shape | Cofactors, steroids, fatty acid sites, catalytic triad in proteases |
2004 | SitesBase [111] | Ligand | Match of triplets of points | Count of matches, RMSD | Cofactors, phosphate sites |
2007 | Ramensky et al. [112] | Ligand | Clique detection in graph of atoms | Dice similarity of matches | Diverse |
2008 | Binkowski et al. [113] | CAST [6] Ligand |
Comparison of pairwise distance histograms | Kolmogorov–Smirnov divergence, overlap of volume, RMSD | Cofactor sites, HIV proteases |
2008 | PocketMatch [19] | Ligand | Comparison of sorted pairwise distances | Normalized count of matches | Diverse, SCOP classes |
2008 | SiteAlign [20] | Ligand | Alignment of polyhedron fingerprints | Normalized distances of fingerprints | Proteases, kinases, estrogen receptors, GPCRs |
2008 | SOIPPA [114] | Ligand | Clique detection in graphs of atoms |
Composite weighted by frequencies, PSSM, distances | Cofactor sites, SCOP classes |
2009 | SMAP [33] | Ligand | Clique detection in graphs of atoms |
Gaussian densities from distances, angles of normal vectors, BLOSSUM weights | Cofactor sites |
2010 | BSSF [25] | PASS [64] | Comparison of fingerprints of binned distances and properties | Canberra distances of fingerprints | Diverse, synthetic data, SCOP classes |
2010 | Feldman et al. [30] | Ligand | Match of subsets of Cα atoms | Probabilistic score from distances between matches | Diverse, kinases |
2010 | FuzCav [24] | Ligand | Fingerprints of triplets of atom features | Maximal proportion of matches | Diverse, functional groups, 8 difficult cases |
2010 | Milletti et al. [115] | Ligand | Comparison of 3 concentric spheres fingerprints encoding neighborhood for each point, solving linear assignment | Composite of fingerprint distances and RMSD | ATP sites, kinases |
2010 | P.A.R.I.S (sup-CK) [116] |
Ligand | Initial alignment optimized by gradient ascent to maximize a Gaussian kernel | Gaussian kernel | Cofactor sites |
2010 | ProBiS [31] | Ligand | Maximum clique detection in graphs of surface atoms | Count of Matches, RMSD, angle between vectors | Cofactor/metal sites, protein–protein interfaces, protein–DNA complexes |
2011 | PocketAlign [117] | Ligand | Initial pairs from sorted lists of atom distances, then extend | Count of matches, RMSD | Cofactor sites, SCOP classes |
2011 | PocketFEATURE [118] | Ligand | Comparison of 7 concentric spheres fingerprints encoding neighborhood for each microenvironment | Normalized Tanimoto similarity of fingerprints | Kinases |
2012 | KRIPO [21] | Ligand | Fingerprints of triplets of pharmacophore | Modified Tanimoto of fingerprints | Diverse, fragments subpockets, search of bioisosteric substructures |
2012 | Patch-Surfer [119] | Ligand LIGSITE [12] |
Comparison of 3D Zernike of surface patches solving a weighted bipartite matching |
Composite of surface match distances and size differences | Cofactor sites |
2012 | Shaper [47] | VolSite [47] | Comparison of cloud of points by Gaussian shapes matching | Tanimoto, Tversky of matches | Diverse, GPCRs, proteases |
2012 | TIPSA [120] | Ligand | Match of quadruplets of points, iterative refinement by Hungarian algorithm |
Tanimoto of matches, overlap of volume, normalized RMSD | Cofactor sites |
2013 | Apoc [28] | Ligand CAVITATOR [28], LIGSITE [12] |
Seed alignment by comparing secondary structures, optimized by solving linear assignment problem | Composite of vector orientation, distance, properties | Diverse, similar ligand recognition sites |
2013 | TrixP [121] | DoGSite [122] | Search for common shape and triplets of points by bitmap indexing | Composite of matches count, angle between vectors, mismatches penalty | Diverse, 8 difficult cases, protease, estrogen receptor, HIV reverse transcriptase |
2014 | eMatchSite [29] | eFindSite [123] | Template-based alignment optimized by Hungarian algorithm | Machine learning score: RMSD, residue, properties | Cofactors, steroid sites |
2014 | RAPMAD [26] | LIGSITE [12] | Comparison of 14 pairwise distance histograms, one for each property |
Jensen–Shannon divergence of histograms | Cofactor sites, proteases, diverse |
2015 | IsoMIF [124] | GetCleft [125] | Clique detection in graphs of interaction grid points |
Tanimoto of descriptors of matched points | Cofactors, steroid sites |
2016 | G-LoSA [126] | Ligand | Clique detection in graphs of atoms |
Feature-weighted count of matches | Diverse, Ca+ sites, similar ligands recognition sites, protein–protein interfaces |
2016 | SiteHopper [127] | Ligand | Comparison of surface atoms by Gaussian shapes matching | Weighted combination of shape and color Tanimoto | Diverse using binding affinities |
2019 | DeepDrug3D [128] | Ligand | Convolutional neural network model | Binary classification | Cofactors, steroids sites, proteases |
2020 | DeeplyTough [32] | Fpocket [16] Ligand |
Convolutional neural network model | Binary classification | Cofactor sites, diverse and using binding affinities |
2020 | ProCare [129] | VolSite [47] | Match of randomly sampled quadruplets refined by iterative closest point | Tversky of matched pharmacophoric properties |
Diverse, using functional annotation, fragments subpockets, search of bioisosteric structures |
2021 | PocketShape [130] | Ligand | Initial alignment optimized by Hungarian algorithm | Composite of matches, orientation of residues | Diverse SCOP classes, kinases |
2021 | Site2Vec [131] | Ligand | Random forest model on autoencoder-generated descriptors | Binary classification | Cofactors, steroid sites, diverse |