NETBAG: An efficient and accurate gene-network prediction tool
Mutations in DNA are fundamental to evolution, adaptation and the great diversity among species. They are also the basis of genetic disease. Investigating genetic mutations in complex human phenotypes such as autism, schizophrenia, or autoimmune disorders continues to be a major challenge in biology, as it is a time and labor intensive process. Further complicating these studies is that individual mutations alone can be benign, but combinations of these mutations lead to disease. This technology, termed NETBAG (network-based analysis of genetic associations), is modeling software that identifies networks of gene interactions underlying complex human phenotypes. Interaction networks have proven extremely useful in recognizing the molecular underpinnings of complex diseases and may provide a cohesive mechanism for identifying mutations in seemingly unrelated genes. Furthermore, identifying the most likely nodes for phenotypic networks allows researchers to focus their efforts on confirming the relevance of the most likely candidates, greatly reducing the time and monetary costs associated with validating potential drug targets.
Network and cluster analysis can predict phenotypic consequences of seemingly unrelated genes
This technology takes a list of genes, most likely identified from patient probands, and utilizes information from numerous sources to recognize and score the probability of network interactions under a unified statistical framework. Information proven useful in network recognition includes physical protein-protein interactions, gene expression profiles, and shared interaction partners and annotations. This information is passed through an algorithm employing naive Bayesian integration and comparative genomics from model organisms to determine network significance. Beyond identifying novel targets for a disease, the technology may also prove extremely valuable to companies who have already developed a treatment targeted at a specific gene and wish to expand its use to other diseases that also contain that gene within their predicted network.
The technology has been rigorously validated using publicly available databases to identify previously confirmed network interactions and proof-of-concept experiments have been performed to identify novel networks involved in autism and schizophrenia phenotypes.
Lead Inventor:
Applications:
- Genetic disease research and modeling
- Bioinformatics
- Drug discovery
- Family planning
- Genetic counseling
- Personalized medicine
- Independent validation of genome wide association studies (GWAS) or other whole genome technologies such as whole genome sequencing, next-gen sequencing, or exome sequencing
- Identifying diseases that may be treated with an already approved medication originally developed for a gene that is common between the disease networks
Advantages:
- Integration of diverse datasets increases the probability of identifying subtle or complex interactions
- Can discern which types of mutations (e.g. nonsense, missense, copy number variant, etc.) may be more likely to contribute to disease phenotypes
- Utilizes publicly available databases that are continuously curated and updated for employing the most recent advances and discoveries
- Can stratify individual patients with a common phenotype based on their individual network perturbations to customize diagnostic, prognostic, and therapeutic predictions
Tech Ventures Reference: IR CU14362