EVENT : C3BI Seminars
Chemoinformatics and proteochemometrics approaches to the pharmacological modulation of protein-protein interactions.
Main speaker : Olivier Sperandio, from Chemoinformatics and Proteochemometrics group, Institut Pasteur, Paris Date : 15/12/2016 at 02:00 pm Location : Retrovirus room – LWOFF (22) ,Institut Pasteur, Paris
Pharmaceutical innovation is still impaired by the paucity of clinically testable targets and by the fact that only a few are successfully exploited in each therapeutic area. This stands in sharp contrast with the number and diversity of roles of Protein-Protein Interactions (PPIs) in cells. Indeed, with about 130,000 binary PPIs and possibly more just in humans, the development of drugs targeting these systems, represents a significant step toward expanding the druggable genome and a possible leverage on the pharmacological modulation of disease-associated cellular pathways. Yet, despite the importance of PPI in biology, this class of targets has been extremely challenging to translate into therapeutics. This is explained by the features of the PPI interfaces themselves including relatively large contact surface areas and the lack of clearly identifiable grooves or deep pockets hindering the effective binding of small molecules. The key of success in finding chemical probes for PPI targets is two fold. First, which PPI target should be selected for such purpose as we now know that not all PPIs are suited for pharmacological modulation? Second, once the PPI target has been selected, which chemotypes should be used to address this target as it is well documented that commercial chemical libraries are poorly suited for such an endeavor. Lessons from the past have taught the community that, although challenging, designing chemical probes and drugs targeting PPI can be feasible when the right combination of technologies is used. In this presentation I will discuss some of the approaches we use to address this challenge using chemoinformatics, structural bioinformatics, and proteochemometrics. At the core of our strategy is the gathering of successful examples of PPI modulation to improve our way to design starting-high quality PPI probes. In 2011, we decided to develop iPPI-DB (http://www.ippidb.cdithem.fr), a manually curated database from the scientific literature that contains the structure, some physicochemical characteristics, the pharmacological data and the profile of the PPI targets of several hundred modulators of protein-protein interactions. It also provides information about similarity with known drugs, and links to external databases like PubChem, and ChEMBL. It has been developed as a tool to assist the chemists and biologists in the identification of chemical starting points to study a PPI of interest. We use such data to build machine learning models in order to identify privileged chemical substructures that may serve as quality chemical starting points to a wide range of PPI targets. Conversely, we are interested in the structural data of the PPI targets themselves and use structural bioinformatics to scrutinize binding cavities at the core of interfaces. Such procedures help us to prioritize targets and adapted chemotypes to modulate them.
Due to security policy in Institut Pasteur, please register before if you plan to come to this meeting