Cambridge, Mass.—Gene Network Sciences Inc. (GNS) has entered into a research collaboration with the University of California San Francisco Cancer Center (UCSF) aimed at accelerating cancer research and drug development across several therapeutic areas. This collaboration will combine the clinical and research oncology expertise of UCSF with the computational expertise and supercomputer-driven REFS (reverse engineering and forward simulation) platform of GNS. Financial terms of the agreement were not disclosed.
"GNS has a unique computational approach to analysis," says Dr. W. Michael Korn, associate professor at UCSF and leader of the UCSF-GNS research team, "with the capability to handle large scale data of cell change in silico and answer the question, 'What would happen if?' using Bayesian network inference modeling."
A Bayesian network is a probabilistic graphical model that represents a set of random variables and their conditional independencies. These models are then simulated to discover both new targets for drug intervention and genetic markers of drug response that match patients who will respond to a given drug treatment with a particular clinical trial and treatment option.
By discovering how and why specific sets of genes and drug candidates impact human biology, GNS technology aims to promote the rapid development of breakthrough drug and diagnostic products and the matching of patients to the optimal therapy.
GNS was incorporated in 2000 and moved to Cambridge three years later, says Dr. Iya Khalil, executive vice president and co-founder of the company. "Our focus has been on generating a platform that can translate genomic, proteomic, metabolomic and phenotypic data into models of human disease."
"Both parties bring their unique expertise to the challenge," notes Tom Neyarapally, senior VP of corporate development at GNS. "UCSF has the biologists, patients and data. We have the supercomputer expertise." The two organizations will share in any commercial benefits of their joint venture, Neyarapally adds.
GNS and UCSF have utilized REFS to discover and validate novel mediators of the cell cycle transition, which is a critical determinant of the rate of cancer cell proliferation and tumor growth. These results were presented by UCSF researcher Dr. Rina Gendelman, a postdoctoral fellow in Korn's laboratory, at the American Association for Cancer Research (AACR) 2009 meeting in Denver, during which Dr. Gendelman received an AACR-Susan G. Komen Scholar-in-Training Award. UCSF is one of nine centers in the National Institutes of Health (NIH) Integrated Cancer Biology Program headed by UC-Berkeley's Dr. Joe Gray.
Khalil notes that GNS has developed novel targets and compounds that are moving into the clinic and estimates that perhaps 30 percent of new drugs aimed at combating diabetes are traveling down the in silico discovery pathway. In addition, she notes that the same kind of data can be used in patient care as electronic medical records become a reality over the next five years. Dr. Korn adds that cancer is "in most cases a conglomerate of complex changes. Linear connections don't tell the real story," he notes. "Our approach can solve the mysteries."
"Personalized medicine can mean a lot of different things," Neyarapally notes. "Many of us can't hold more than five variables in our minds. TV's Dr. House can handle 15 or 20—definitely not 100,000. Supercomputers can plow through terabyte after terabyte of data. The models that we're building will be able to assist MDs in making decisions—providing a high-powered PDR to ask 'what if' questions. The model will become rich enough to provide more predictive, more effective answers."
The parties will continue with research to elucidate the RAS-MAPK and PI3K cancer pathways that are critical for initiation and progression of many cancers by building models of breast, pancreatic and esophageal cancer based on data from ongoing research at UCSF using the REFS platform and simulation of these models. The discoveries from the ongoing research are expected to include potential novel drug targets in those cancer types.
"GNS has a unique computational approach to analysis," says Dr. W. Michael Korn, associate professor at UCSF and leader of the UCSF-GNS research team, "with the capability to handle large scale data of cell change in silico and answer the question, 'What would happen if?' using Bayesian network inference modeling."
A Bayesian network is a probabilistic graphical model that represents a set of random variables and their conditional independencies. These models are then simulated to discover both new targets for drug intervention and genetic markers of drug response that match patients who will respond to a given drug treatment with a particular clinical trial and treatment option.
By discovering how and why specific sets of genes and drug candidates impact human biology, GNS technology aims to promote the rapid development of breakthrough drug and diagnostic products and the matching of patients to the optimal therapy.
GNS was incorporated in 2000 and moved to Cambridge three years later, says Dr. Iya Khalil, executive vice president and co-founder of the company. "Our focus has been on generating a platform that can translate genomic, proteomic, metabolomic and phenotypic data into models of human disease."
"Both parties bring their unique expertise to the challenge," notes Tom Neyarapally, senior VP of corporate development at GNS. "UCSF has the biologists, patients and data. We have the supercomputer expertise." The two organizations will share in any commercial benefits of their joint venture, Neyarapally adds.
GNS and UCSF have utilized REFS to discover and validate novel mediators of the cell cycle transition, which is a critical determinant of the rate of cancer cell proliferation and tumor growth. These results were presented by UCSF researcher Dr. Rina Gendelman, a postdoctoral fellow in Korn's laboratory, at the American Association for Cancer Research (AACR) 2009 meeting in Denver, during which Dr. Gendelman received an AACR-Susan G. Komen Scholar-in-Training Award. UCSF is one of nine centers in the National Institutes of Health (NIH) Integrated Cancer Biology Program headed by UC-Berkeley's Dr. Joe Gray.
Khalil notes that GNS has developed novel targets and compounds that are moving into the clinic and estimates that perhaps 30 percent of new drugs aimed at combating diabetes are traveling down the in silico discovery pathway. In addition, she notes that the same kind of data can be used in patient care as electronic medical records become a reality over the next five years. Dr. Korn adds that cancer is "in most cases a conglomerate of complex changes. Linear connections don't tell the real story," he notes. "Our approach can solve the mysteries."
"Personalized medicine can mean a lot of different things," Neyarapally notes. "Many of us can't hold more than five variables in our minds. TV's Dr. House can handle 15 or 20—definitely not 100,000. Supercomputers can plow through terabyte after terabyte of data. The models that we're building will be able to assist MDs in making decisions—providing a high-powered PDR to ask 'what if' questions. The model will become rich enough to provide more predictive, more effective answers."
The parties will continue with research to elucidate the RAS-MAPK and PI3K cancer pathways that are critical for initiation and progression of many cancers by building models of breast, pancreatic and esophageal cancer based on data from ongoing research at UCSF using the REFS platform and simulation of these models. The discoveries from the ongoing research are expected to include potential novel drug targets in those cancer types.
