2D- and 3D-Similarity based classification systems for substrate prediction of ABCB1

COMP 209

Rita Schwaha, rita.schwaha@univie.ac.at and Gerhard F. Ecker, gerhard.f.ecker@univie.ac.at. Department of Pharmaceutical Chemistry, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria
As part of the ATP-binding cassette transporter superfamily ABCB1 (P-gp) exports a multitude of xenobiotics and is strongly connected to multi-drug resistance (MDR). A valid crystal structure of this protein is still missing and ligand based approaches are the methods of choice. Its high promiscuity and implications in drug/drug interactions and brain permeation of drugs renders prediction of possible substrates highly demanding. Recently we could demonstrate that Similarity Based Descriptors (SIBAR) are a useful tool to establish predictive in silico models for drug/ABCB1 interaction. Based on a reference set representing satellite structures in the chemical space we compared a 3D-shape based comparison (ROCS) with SIBAR based on VSA and Volsurf descriptors. Our results validated on an external test set showed an overall prediction accuracy of 74 % with the VolSurf Descriptors achieving the best overall classification performance.

Financial support provided by the Austrian Science fund (L344-N17 and F3502).

 

Poster Session
6:00 PM-8:00 PM, Tuesday, August 18, 2009 Walter E. Washington Convention Center -- Ballroom A, Poster

Sci-Mix
8:00 PM-10:00 PM, Monday, August 17, 2009 Walter E. Washington Convention Center -- Hall D, Sci-Mix

Division of Computers in Chemistry

The 238th ACS National Meeting, Washington, DC, August 16-20, 2009