Table 2 Datasets Used to Evaluate Predictors in Table 4 , including the source from which they were derived, as well as the publication in which they were created using the requirements in the “Description” column
From: Algorithmic approaches to protein-protein interaction site prediction
| Label | Name | Derived from | Source | Description | Creator | Year |
|---|---|---|---|---|---|---|
| A | DB3-188 | DB 3.0 | [77] |
\(\mathcal {S}_{i} < 40\% \);  >50
| [31] | 2010 |
| B | DS56B | CAPRI | [78] | Targets 1-27 Bound | [31] | 2010 |
| C | DS56U | CAPRI | [78] | Targets 1-27 Unbound | [31] | 2010 |
| D | NI1 | PDB | [79] |
\(\mathcal {S}_{i} < 70\% \);  ≤3.5; 100≤ ≤800;
| [37] | 2014 |
| N 100≥1; Excl. Ag-Ab; Non-obligate; MC | ||||||
| E | NI2 | PDB | [79] |
\(\mathcal {S}_{i} < 70\% \); ≤3.5; 100≤ ≤800;
| [37] | 2014 |
| N 100≥2; Excl. Ag-Ab; Non-obligate; MC | ||||||
| F | PlaneDimers | Mintz et al. | [110] | Planar PPI; 20%<MSA i <90%; | [33] | 2011 |
|
Excl. MBPs, Ag-Ab, VS; >100; Perm
| ||||||
| * | Dimers | Mintz et al. | [110] | Clustered on seq. similarity; Excl. | [33] | 2011 |
|
MBPs, Ag-Ab, VS; >100; Perm
| ||||||
| G | TransComp_1 | DB 4.0 | [108] | “Simple” (low conf. change); Non-obligate | [33] | 2011 |
| * | TransComp_2 | CAPRI | [111] | Not in TransComp_1; Non-obligate | [33] | 2011 |
| H | W025 | DB 1.0/2.0 | [106,107] | Excl. Ag-Ab, enzyme interactions | [41] | 2006 |
| I | S435 | PDB | [79] |
PQS filtered; \(\mathcal {S}_{i} < 50\% \); >30;
| [39] | 2007 |
| Excl. NA, MBPs, VS, NMR | ||||||
| J | S149 | PDB | [79] |
PQS filtered; \(\mathcal {S}_{i} < 50\% \); >30; Excl.
| [39] | 2007 |
| NA, MBPs, VS, NMR; N H (S435) | ||||||
| * | S21a | S149 | [39] | Nonredundant; MC | [39] | 2007 |
| K | S58 | PDB | [79] |
\(\mathcal {S}_{i} < 30\% \); ≤3.0; >100;
| [7] | 2012 |
| Excl. NA, ligands; N H (S435) | ||||||
| L | 3DS | 3did | [112] |
\(\mathcal {S}_{i} < 25\% \); >50
| [38] | 2012 |
| M | B100 | DB 3.0 | [77] | Excl. Ag-Ab | [40] | 2011 |
| N | BM180 | PDB | [79] |
\(\mathcal {S}_{i} < 20\% \); ≤3.0; >20; Excl.
| [13] | 2005 |
| NMR; Divided into 4 sub-types | ||||||
| * | S1 | PDB | [79] |
\(\mathcal {S}_{i} < 50\% \); 10< <30; Disordered
| [113] | 2009 |
| short; Excl. MBPs, NA; Disprot filtered | ||||||
| * | S2 | PDB | [79] |
\(\mathcal {S}_{i} < 50\% \); >30; Disordered
| [113] | 2009 |
| long; Excl. MBPs, NA; Disprot filtered | ||||||
| * | DS24Carl | PDB | [79] |
>20; 8 Perm + 16 Non-obligate
| [66] | 2008 |
- The “Label” column defines the alphabetic character used to refer to the dataset in Table 4. “ ∗” in the “Label” column signifies that the set is not presented in Table 4 as it is not widely used. \(\mathcal {S}_{i}\) is the sequence identity redundancy cutoff,
is the amino acid length of the chain, 
is the resolution cutoff in angstroms, N
100≥n requires that the number of interface residues per 100 residues in a given protein to be greater than n, Ag-Ab refers to antigen-antibody complexes, M
S
A
i
is the sequence identity redundancy cutoff for chains in an MSA, VS refers to Viral Subunits, NA refers to Nucleic Acids, N
H
(x) refers to a set being non-homologous to set x, MC denotes that both the monomer and the complex to which it belongs are known.