Novel software for predicting unknown drug-target interactions
- Detailed Technology Description
- None
- *Abstract
-
The invention is a software package that predicts new (unknown) associations between known drugs and their targets (usually proteins), thus allowing the identification of new potential drugs, among those already FDA-approved, to target well-defined proteins or pathways. Application1. By predicting unknown associations, this software can solve problems such as assessing the mechanism through which a drug produces its side effects, the therapeutic activity of a drug in greater detail and, given some information about a drug lead, the software can predict potential side effects or toxicity/lethality. 2, Determines unknown interactions of a drug-target pair by reducing the number of possible interactions that can be tested from millions to a handful, using biological information. Advantages1. First known method that combines biological information about targets with graphical models to perform interaction predictions entirely in silico. 2. Analyzes the entire ligand-target interaction space without the need for any chemical synthesis, protein purification or assay design. 3. Works with real datasets containing more than 1000 drugs and targets * i.e. more than a million possible interactions * .4. Produces its results within minutes. Copyright
- *Principal Investigator
-
Name: Ivet Bahar, Professor and John K Vries Chair
Department: Med-Computational and Systems Biology
Name: Murat Cobanoglu
Department: Med-Computational and Systems Biology
- Country/Region
- USA
For more information, please click Here
