Data on protein-drug interactions are rapidly increasing and being collected in databases such as DrugBank. The information in these databases usually reflects known/observed interactions, while the lack of data for a given protein-drug pair does not necessarily mean that those protein-drug molecules are not interacting. Indeed, recent studies supported by both computations and experiments indicate that many drugs have side effects (i.e. they target proteins) other than those known/compiled in DrugBank. This latter property may be exploited for designing 'repurposable' drugs or polypharmacological treatments. Efficient identification of such potential interactions is an important challenge that is likely to accelerate drug discovery and development efforts. There is a need for efficient identification of such data, and efficient dissemination of results.
BalestraWeb addresses this need by learning a probabilistic latent factor model of drug-target interactions and enabling users to predict the interactions of:
- any known drug against all targets (by providing only a drug identifier)
- any target against all known, approved drugs (by providing only a target identifier)
- any drug-target pair (by providing both a drug and a target identifier)
Built by: Murat Can Cobanoglu.
BalestraWeb 2.0 is re-built by Hongchun Li using PHP, Python and PostgreSQL database instead of the Python web framework. The results on this server were generated by Hongchun Li and Fen Pei using the same method described in the version 1.0 and the dataset was extended to both the FDA approved and all drugs in the DrugBank version 5. The dataset is also being extended to the STITCH 5.0 for human.
- Bahar Lab at the
University of Pittsburgh School of Medicine,
Dept of Computational and Systems Biology
- University of Pittsburgh Drug Discovery Institute,
Funding: Support for BalestraWeb development is provided by NIH awards U19 AI068021 and P01 DK096990.
Copyright © 2013-2014, University of Pittsburgh