Scoring protein–protein interaction (PPI) models with PrePPI.A, creation of a model (yellow–green) for the interaction between human Cullin-5 (CUL5) and defective in cullin neddylation protein 1 (DCN1)-like protein 5 (DCUN1D5) based on structural alignment of models for CUL5 (yellow) and DCUN1D5 (green) to the PDB template 3o2p:EA complex between the yeast cell division control protein 53 (CDC53; 3o2p, chain E, brown) and DCN1 (3o2p, chain A, purple). In spite of low sequence identities—23% for CUL5 versus CDC53, and 25% for DCUN1D5 versus DCN1—the models for CUL5 and DCUN1D5 align well to the template chains with low protein structure distances (32) of 0.05 and 0.04. The average of the protein structure distance values contributes to the PrePPI structure modeling (SM) score. B, structure superposition of models for CUL5 (yellow) and DCUN1D5 (green) on the PDB template 3o2p:EA. The dotted circle highlights the interaction interface, a portion of which is described in (C). C, structure-based sequence alignments for portions of the interface (dotted circle in (B)): CUL5 (yellow) on CDC53 (3o2p:E, brown), and DCUN1D5 (green) on DCN1 (3o2p:A, purple). The interaction model (yellow–green) is not explicitly scored; rather its relationship to the template (brown–purple) is evaluated for the PrePPI SM score. Positions of residues that make interfacial contacts in 3o2p:EA (i.e., between CDC53 and DCN1) are bolded and boxed. The alignments reveal that all the residues involved in the template interface are aligned to residues in the query proteins, and this yields a favorable contribution to the SM score. Then, the propensity for these model residues to be interfacial is evaluated. PredUs (38) predicts whether residues in CUL5 and DCUN1D5 are interfacial (blue) or not (red). The contacts observed for R804 in the template are shown as lines: The solid lines denote the cases where the contacting template residues are aligned to residues in CUL5 and DCUN1D5 that are predicted to be interfacial (blue dots); the dotted line denotes the case where contacting template residues are aligned to model residues that are not predicted to be interfacial (red dots). Only those contacts that are likely to occur between the proteins in the interaction model (solid lines) further increase the SM score.