|
More than 40% of the US population is affected by some form of cancer in their lifetime. While we acknowledge that the term “cancer” is a “catch-all” for a number of different diseases with varying etiologies, therapeutic options and probable outcomes, there is an underlying complex of related, interacting protein pathways that appear to govern the cellular conversion to a metastatic state. Understanding significant contributions to the web of interactions, combined with early detection and appropriate therapy will be critical to improving survival rates and developing useful diagnostic and therapeutic approaches.
UCSD inventors have developed an approach based on the reality that cancer is a ‘disease of interactive pathways' and the assumption that keys for understanding the disease are encoded in protein networks. Their model has been applied toward two endpoints: (1) evaluate the contribution of various genes in a common pathway and the convergence of alternative pathways to the same predictive endpoint and (2) assesses interactive protein pathways (vs. individual genes) to determine the likelihood of metastatic progression once a woman is diagnosed with breast cancer. The ability to pinpoint most-likely therapeutic targets in relevant pathways may also yield insights on causation.
The current technology was developed and validated for improving the prognosis for breast cancer patients. However, it has other, significant applications and differentiating features:
- It is a dynamic model which is able to iterate and self-refine to identify informative modules (e.g., learn what is and is not important)
- It can highlight which genes are likely causes vs. effects of cancer; this feature may identify new therapeutic targets
- The approach may dramatically improve outcomes for profile-based diagnosis. While demonstrated for microarrays, it is readily adapted to classify other datasets, including protein expression, protein phosphorylation and/or metabolic profiles.
- The algorithms may be adapted to identify pathways and mechanisms for other conditions.
- Although demonstrated for two outcomes (metastatic or not), the approach can be expanded to evaluate numerous outcomes.
Perhaps most interestingly, by intelligently targeting the appropriate pathways, this network/module approach may greatly improve the efficacy of combination therapies and profile-based diagnosis.
References and Supplemental Information:
Ideker Lab
UCSD News Release
Case No: SD2006-131
Keywords: cancer, diagnosis, diagnostic, therapeutic, model. software, pathways, protein, network, module, target, breast
Inquiries To: invent@ucsd.edu
|
