Intellectual Property Office
Non-Confidential Disclosures
" Method for Inferring Genetic Interaction Networks from Gene Expression Patterns "
PSU Invention Disclosure Number 3196
Download a PDF of this description
Field of the Invention:
Genetics and Molecular Biology
Inventors:
Jayanth R. Banavar, Marek Cieplak, Nina Fedoroff, Timothy R. Lezon, Amos Maritan
Background:
Microarray technology permits the simultaneous measurement of the expression levels of all of the genes within a tissue, culture or an organism. Techniques for extracting biological information from microarray data range from simply clustering the genes based on the similarities of their expression profiles to highly contrived schemes that infer networks of gene interactions based on multiple assumptions about the structure of such networks. Many of the conventional techniques for gene network inference rely on empirically-determined user-defined parameters that add to their systematic errors. Few can be readily scaled to the level of the full genome, and inference techniques are generally limited to small subsets of genes.
Invention description:
The disclosed invention is a method for inferring gene interaction networks based on the principle of entropy maximization. It enables the direct calculation of the most probable network of effective gene interactions that are consistent with microarray data without including any external parameters or assumptions. The method uses a systems approach and is computationally inexpensive enough to be scaled to the level of full genomes. It may also be used to calculate three-gene and higher-order interactions.
Advantages:
- Scalable
- Fast
- Easily incorporated into existing microarray analysis software packages
- No free parameters
- Fully automated
Patent status:
Patent pending.
Contact:
James F. Kolonay
Intellectual Property Office
113 Technology Center
The Pennsylvania State Univ.
University Park, PA 16802-7000
Phone: (814) 863-7070
Fax: (814) 865-3591
E-mail:jfk11@psu.edu
|
