2.1 Benchmarking dataset
Benchmarking dataset of protein-protein and protein-peptide complexes were taken from Vrevenet al. [39] and Hauser et al. [40], respectively. These datasets were earlier used for benchmarking studies of docking programs. Protein-protein benchmarking dataset is composed of non-redundant high-quality structures (resolution < 3.25 Å). Protein−peptide complexes were composed of peptides with 3 to 12 residues (resolution < 2.0 Å, R-free < 0.3). The dataset consists of 134 protein-protein and 50 protein-peptide complexes. The PDB IDs are provided in Table 1.