|
| Reference |
CI techniques | Binding sites/residues | Performance | Datasets |
|
| [23] | ANN | DNA | Accuracy: 64%,
sensitivity: 69% | Amino acid sequence composition, solvent accessibility, and secondary structure |
|
| [24] | ANN | DNA | Accuracy: 73.6% | Position specific scoring matrices (PSSM) |
|
| [25] | SVM | DNA | Accuracy: 90% | Surface and overall composition, overall charge and positive potential patches on the protein surface |
|
| [26] | SVM | DNA | Accuracy: 82.30% | Amino acid sequence, PSSM, and low-resolution structural information |
|
| [18] | SVM | DNA | Accuracy: 77.2%, sensitivity: 76.4%, and specificity: 76.6% | Position specific scoring matrices (PSSM) |
|
| [27] | SVM | DNA | Accuracy: 96.6%, sensitivity: 90.7% | Amino acid sequence, pseudoamino acid composition, autocross-covariance transforms, and dipeptide composition |
|
| [28] | SVM | DNA | Accuracy: 80%, sensitivity: 85.1%, and specificity: 85.3% | Normalized PSSM score, normalized solvent accessible surface area, and protein backbone structure |
|
| [29] | SVM | DNA | MCC: 0.67, accuracy: 89.6%, sensitivity: 88.4%, and specificity: 90.8% | PSSM, amino acid composition, hydrophobicity, polarity, polarizability, secondary structure, solvent accessibility, normalized Vander Waals volume, binding propensity, and nonbinding propensity |
|
| [30] | Ensemble of ANN and SVM | DNA | Accuracy: 89.00% | PSSM and structural features such as secondary structure, solvent accessibility, and globularity |
|
| [31] | Random forest | DNA | Accuracy: 78.20%, sensitivity: 78.06%, and specificity: 78.22% | PSSM with mean and standard of deviation side chain pKa value, hydrophobicity index, and molecular mass |
|
| [32] | Random forest | DNA | Accuracy: 91.41%, MCC: 0.70, and AUC: 0.913 | PSSM, secondary structure information, and orthogonal binary vector information and two physical-chemical properties dipoles and volumes of the side chains |
|
| [33] | Random forest | DNA | Accuracy: 83.96% | Pseudoamino acid composition |
|
| [34] | Gaussian Naive Bayes | DNA | Accuracy: 79.10% and MCC: 0.583 | PSSM, predicted secondary structure, predicted relative solvent accessibility |
|
| [35] | Naive Bayes classifier | RNA | Accuracy: 85.00% | Amino acid sequence, relative accessible surface area, sequence entropy, hydrophobicity, secondary structure, and electrostatic potential |
|
| [36] | SVM | RNA | MCC: 0.31 | Amino acid sequence and PSSM |
|
| [37] | SVM | RNA | Accuracy: 87.99%, sensitivity: 79.95%, and specificity: 90.36% | Smoothed PSSM with the correlation and dependency from the neighboring residues |
|
| [38] | SVM | RNA | AUC: 0.83 | PSSM, accessible surface area, between centrality and retention coefficient |
|
| [39] | Random forest | RNA | MCC: 0.5637, accuracy: 88.63%, sensitivity: 53.70%, and specificity: 96.97% | PSSM, physicochemical properties of amino acids, polarity-charge, and hydrophobicity |
|
| [40] | SVM | RNA | Accuracy: 79.72% and
MCC: 0.59 | Protein sequence, amino acid composition, hydrophobicity, secondary structure, predicted solvent accessibility, normalized Vander Waals volume, polarity, and polarizability |
|
| [41] | SVM | rRNA, RNA, and DNA | rRNA accuracy: 84%,
RNA accuracy: 78%,
DNA accuracy: 72% | Protein sequence, amino acid composition, hydrophobicity, secondary structure, predicted solvent accessibility, normalized Vander Waals volume, polarity, and polarizability |
|
| [19] | SVM | DNA and RNA | DNA sensitivity: 69.40%, specificity: 70.47%, and RNA sensitivity: 66.28%, and specificity: 69.84% | Side chain pKa value, hydrophobicity index and molecular mass |
|
| [42] | SVM | DNA and RNA | Accuracy: 79.00%, sensitivity: 77.30%, specificity: 79.30% for DNA, and accuracy: 77.70%, sensitivity: 71.60%, and specificity: 78.70% for RNA-binding residues | PSSM with mean and standard of deviation side chain pKa value, hydrophobicity index, and molecular mass |
|
| [43] | SVM | Metal binding | Accuracy: 78.10% | Physiochemical properties of the amino acid sequences |
|
| [44] | Bayesian classifier | Zinc | Specificity: 99.8%, sensitivity: 75.5% | Structural properties of a protein |
|
| [45] | Structural comparison | DNA | Accuracy: 98% and precision: 84% | Combination of structural comparison and the evaluation of statistical potential |
|
| [46] | Structural comparison | RNA | Accuracy: 98%, precision: 91% for predicting RBPs, and accuracy: 93% and precision: 78% for predicting RNA binding residues | Distance-scaled, finite, ideal gas reference based statistical energy function, and structural alignment |
|
