Comparison of ultra‐fast 2D and 3D ligand and target descriptors for side effect prediction and network analysis in polypharmacology
- 17 September 2013
- journal article
- research article
- Published by Wiley in British Journal of Pharmacology
- Vol. 170 (3) , 557-567
Abstract
Background and Purpose: Some existing computational methods are used to infer protein targets of small molecules and can therefore be used to find new targets for existing drugs, with the goals of re‐directing the molecule towards a different therapeutic purpose or explaining off‐target effects due to multiple targeting. Inherent limitations, however, arise from the fact that chemical analogy is calculated on the basis of common frameworks or scaffolds and also because target information is neglected. The method we present addresses these issues by taking into account 3D information from both the ligand and the target.Experimental Approach: ElectroShape is an established method for ultra‐fast comparison of the shapes and charge distributions of ligands that is validated here for prediction of on‐target activities, off‐target profiles and adverse effects of drugs and drug‐like molecules taken from the DrugBank database.Key Results: The method is shown to predict polypharmacology profiles and relate targets from two complementary viewpoints (ligand‐ and target‐based networks).Conclusions and Implications: The open‐access web tool presented here (http://ub.cbm.uam.es/chemogenomics/) allows interactive navigation in a unified ‘pharmacological space’ from the viewpoints of both ligands and targets. It also enables prediction of pharmacological profiles, including likely side effects, for new compounds. We hope this web interface will help many pharmacologists to become aware of this new paradigm (up to now mostly used in the realm of the so‐called ‘chemical biology’) and encourage its use with a view to revealing ‘hidden’ relationships between new and existing compounds and pharmacologically relevant targets.Keywords
This publication has 45 references indexed in Scilit:
- Automated design of ligands to polypharmacological profilesNature, 2012
- Large-scale prediction and testing of drug activity on side-effect targetsNature, 2012
- Chemical Structural Novelty: On-Targets and Off-TargetsJournal of Medicinal Chemistry, 2011
- Cytoscape 2.8: new features for data integration and network visualizationBioinformatics, 2010
- Conformer Generation with OMEGA: Algorithm and Validation Using High Quality Structures from the Protein Databank and Cambridge Structural DatabaseJournal of Chemical Information and Modeling, 2010
- Targeting the cancer kinome through polypharmacologyNature Reviews Cancer, 2010
- Network analyses in systems pharmacologyBioinformatics, 2009
- Quantifying the Relationships among Drug ClassesJournal of Chemical Information and Modeling, 2008
- Benchmarking Sets for Molecular DockingJournal of Medicinal Chemistry, 2006
- TarFisDock: a web server for identifying drug targets with docking approachNucleic Acids Research, 2006