Structural and Physico-Chemical Alignment of Protein-Protein Interfaces [About I2I-SiteEngine][ Server Help][ Download Software][Home]
Interface-to-Interface (I2I)-SiteEngine compares pairs of interacting protein binding sites.

Interface-to-Interface (I2I)-SiteEngine: a method for structural alignment between two protein-protein interfacesprotein-protein interface

Abstract:
Protein-protein interfaces, which are regions of interaction between two protein molecules, contain the information regarding patterns of interacting functional groups.  I2I-SiteEngine simultaneously aligns between two pairs of binding sites that constitute an interface.  I2I-SiteEngine is based on recognition of similarity of physico-chemical properties and shapes. Therefore it can recognize similar interfaces shared by evolutionary unrelated proteins with totally different sequences and folds.

Short Presentation: with graphical explanations about the method and its applications.

Scoring Functions: the technical details of implementation.

Reference:
Shulman-Peleg A., Mintz S., Nussinov R, Wolfson HJ. Protein-Protein Interfaces: Recognition of Similar Spatial and Chemical Organizations,  In Jonassen and J. Kim (Eds.): Algorithms in Bioinformatics: 4th International Workshop, WABI 2004, Bergen, Norway, 2004.  Springer Verlag (2004), Lecture Notes in Computer Science, Volume 3240, pp. 194-205 [PDF]

Contact: Alexandra Shulman-Peleg  shulmana@post.tau.ac.il

Application: Classification of protein-protein interfaces for recognition of similar spatial and chemical organizations
I2I-SiteEngine was applied to classify a pilot dataset that contains some of the common protein-protein interfaces. As a result, these 64 interfaces were clustered into 22 different groups. Some of the clusters include similar interfaces comprised by proteins with different structural folds.  This may suggest that certain interface binding organizations may be shared by different protein families.  Furthermore, similar interfaces may suggest not only similar binding organizations may also suggest similarity in binding partners and function. 
Click here to view the details of the obtained clusters.
Click here to view the default parameters used in this study.