Protein-protein interactions (PPIs) are capable of highly specific interactions, responsible for a wide range of functions, making them attractive in the pursuit of new therapeutic options. With respect to the interface geometry, early studies focused on the overall geometry of interfaces and concluded that PPI interfaces are generally flat and featureless. However, more recent studies support a view of PPI interfaces wherein small interacting pocket regions and geometry characteristics might play an important role in molecular recognition and complex formation. However, due to the lack of understanding and complexity of PPI’s interface regions, this still remains a challenging area.
Here we present the preliminary results of a large scale analysis of physicochemical and geometrical properties based on all 3D structures of PPIs available in the Protein Data Bank. A threshold of 95% sequence similarity at the interface level was used to generate a non-redundant dataset of 41,023 structures of pairwise interactions. These interfaces were then clustered into symmetric homopairs, where the chains interact using equivalent residues on each side of the interface, or non-symmetric. Furthermore, interfaces were further classed based on their size, as globular-globular, enzyme:peptide and non-enzyme:peptide interactions.
Our preliminary analysis reveals that most PPIs make use of 3-4 segments at the binding interface continuity, except for interactions involving peptides which, in general, utilise a single interacting segment. In addition, in terms of secondary structures present at PPI interfaces, loop residues represented the majority, while sheet residues were significantly less prevalent in all interfaces. We anticipate our analysis to add great value to the understanding of how PPIs work.