In the very recent years the number of small molecule disrupting protein-protein interactions (PPIs) has grown rapidly partly due to the improvement of in silico drug discovery methods. PPIs are now considered as the next generation of very promising therapeutic targets (Roche & Morelli, 2010).

Although a lot of PPI inhibitors have been discovered, little is known about structural parameters that govern the druggability of protein-protein interactions and about the chemical properties of small molecule inhibitors of PPI. We have analyzed the interface of PPI for which both the protein-protein and protein-inhibitor complexes have been structurally characterized to extract the best descriptors of PPI with a known inhibitor.
The chemical space of PPI is not completely covered due to the limited amount of data currently available; however, the definition of what makes a PPI a potentially druggable target will become more and more reliable, as the number of 3D structures increases.

Different tools have been developed to describe and analyze the PPI chemical space:

  2P2I database

Structural Database of Protein-Protein Complexes and their Inhibitors

2P2IDB is a hand-curated database dedicated to the structure of protein-protein complexes with known small molecule inhibitors. We have gathered from the PDB, 14 PPIs and 71 small molecule inhibitors found at the interface  (Basse et al, in press; Bourgeas et al, 2010). Interfaces in 2P2IDB have been analyzed in terms of geometrical parameters, atom and residue properties, buried accessible surface area and other biophysical parameters and compared to those of representative datasets of heterodimeric complexes. The 14 complexes have been classified into 2 clusters based on the number of segments at the interface.

  2P2I inspector

Protein-protein interface analysis tool

2P2IINSPECTOR is a tool to calculate a series of physical and chemical descriptors that characterize protein–protein interfaces from 3D structures (Basse et al, in press). These 48 parameters include, interface area, gap volume, non-bonded contacts, hydrogen-bonds, residue and atomic composition, number of segments and secondary structure contribution. Users can analyze protein complexes from the PDB using standard four letter accession codes or upload their own PDB files. The analyzed complex can be visualized using a Jmol applet containing a user-friendly menu. This new tool thus provides an efficient way of characterizing protein–protein interfaces.

  2P2I score

Druggability Prediction of Protein-Protein Interfaces

2P2ISCORE is a scoring function to predict the druggability of protein-protein interfaces. The current qualitative version of the tool was built by extracting six interface descriptors that allow a good discrimination between PPIs with known inhibitors and transient heterodimers from the PDB (Basse et al, in press; Bourgeas et al, 2010).

  2P2I hunter

Design of Focussed PPI Chemical Libraries

2P2IHUNTER is a filtering tool to design focused chemical libraries dedicated to the inhibition of protein-protein complexes. It was built using a combination of learning approaches (SVM) and general profiles of PPI inhibitors (‘rule-of-4’) derived from the set of orthosteric PPI inhibitors present in the 2P2I database (Morelli et al, 2011).