MAPPIS: Multiple Alignment of Protein-Protein Interfaces

Association and dissociation of protein molecules are crucial for most of the cellular processes. A protein-protein interface (PPI) is defined by a pair of regions of two interacting protein molecules that are linked by non-covalent bonds. Analysis and classification of PPIs may help to recognize certain interface binding organizations shared by different protein families, which may be important for the formation and stability of protein-protein complexes. These features may constitute targets for drug discovery and assist in predicting side effects. Many functionally similar protein binding sites share no similar sequential patterns, thus, they cannot be recognized by commonly used methods like PROSITE. A sequence order independent structural alignment method was used by Keskin et al.[2] to classify all known PPIs according to Cα patterns. However, side chains play an important role in the interactions between molecules. Thus, a more accurate, atomic-level, analysis is required[4]. Recently, we have developed a method for alignment between a pair of PPIs[6]. We have applied it to classify[3] the interfaces of all the protein-protein complexes currently available in the PDB. Here we present a novel method for multiple structural alignment of a set of PPIs. To the best of our knowledge this is the first method that can align and detect common patterns of interaction shared by a set of PPIs. Our main motivation is similar to the multiple sequence alignment thesis, a feature common to a number of proteins is probably functionally more significant than a similar feature found only between a pair of proteins. We present the method and show its application to the dataset created by Mintz et al.[3].