Streamlining Utah's Population Database

This month marks the 30th anniversary of the Utah Population Database (UPDB), an extensive compilation of genetic, epidemiological, demographic, and public health statistics on Utah families compiled by the University of Utah.

Seeded with genealogical records of the Church of Jesus Christ of Latter Day Saints, this repository has allowed researchers at the university and at medical institutions around the world to profile a uniquely homogenous population that remains largely descended from the Mormon families that settled in the region 150 years ago. With its insight into families with higher than normal incidences of cancer or other diseases, the database is regarded as a gold mine for genetics and genomics researchers.

Since 2002, drug and biopharmaceutical firms have been culling UPDB in partnership with LineaGen Research, a nonprofit corporation spun off that year from the University of Utah's Huntsman Cancer Institute, which houses the database. LineaGen, which facilitates access to UPDB by commercial partners, has entered research agreements with Celera Diagnostics, Amgen, Battelle Memorial Institute, and Merck. These partnerships are putting UPDB data to work identifying biomarkers, validating therapeutic targets, and investigating therapies for respiratory and other diseases.

Earlier this year, LineaGen formed an information technology (IT) collaboration with IBM that will impact all of its partnerships. Under the agreement, IBM's health care and life sciences division will use its Clinical Genomics Solution, an IT network based on open standards and research protocols, to develop a streamlined means of querying data from UPDB.

Executives with LineaGen's research partners say a standardized IT infrastructure will expedite genomics research in general, potentially offering an alternative to the ad hoc, laboratory-specific computing rigs that are now common in the field. Standardized IT, they say, will go a long way toward improving the efficiency with which commercial research organizations can navigate the database.

Improved efficiency fits well with the objectives of LineaGen, which, according to Chief Executive Officer Stephen M. Prescott, constitutes the first significant effort to broadly commercialize the data in UPDB. Launched with a $4 million grant from the Utah state legislature-secured with the help of former governor Michael Leavitt, the current secretary of the U.S. Department of Health & Human Services-LineaGen specializes in commercial project management and collaborative research.

LineaGen's partners say they are interested in the Utah database primarily because of its genetic data on families that stem from the uniquely insular history of Utah and efforts, initially on the part of the Mormon church, to keep meticulous records on lineages.

Prescott, who until this year also headed the Huntsman Cancer Institute, notes that the Utah Genetic Reference Project Database, a subset of UPDB composed of data on three generations of 47 founding families, is a kind of Rosetta stone in genomics research. These families are the original CEPH families studied by the Human Genome Project at the Centre d'Etude du Polymorphisme Humain in Paris.

By using UPDB, Prescott says, researchers can take the current population and cluster it into related family groups, on which there is exhaustive information. In the last two years, we've expanded the database tremendously beyond the genealogical records that we got from the church, he says. In partnership with the state, the university has added to the database all of the state's vital records going back to the early part of the 20th century-birth certificates, death certificates, marriage licenses, and even driver's licenses-creating a demographic record of nearly every resident of Utah.

According to Prescott, 6070% of the people in the state are somehow related to one of the founding families. With the data available, researchers can perform control studies of, for example, everybody with diabetes. Or they can conduct a typical genetics study. Some families have thousands of living members descending from one of the original founders, Prescott says. Genetically, that is incredibly powerful. In my opinion, studies using families are the most powerful type when it comes to this gene causes that.' And nobody else has a database like this.

Mapping Lancaster

Microarrays Speed A Small Clinic's Research

Scientific interest in the genetic makeup of isolated human populations spiked with the decoding of the human genome. The Utah Population Database and deCode Genetics ' study of the genetic makeup of the population of Iceland are two high-profile efforts to unveil disease mechanisms by plumbing significant eddies in the global gene pool. A lesser known effort is under way in eastern Pennsylvania.

For 17 years, the Clinic for Special Children, a private pediatric health care facility nestled amid the cornfields of Lancaster County, Pa., has been caring for local children with genetic disorders. Most patients are members of the Mennonite and Amish religious sects, closed communities that migrated to the region from Alsace Lorraine 300 years ago.

D. Holmes Morton, founder and medical director of the clinic, says the facility has worked over the years to identify and diagnose rare diseases prominent among the local population-a tall order for a community clinic that is still financed partly by quilt sales. The facility stepped up its efforts in 1998 with the hiring of Erik G. Puffenberger, a molecular biologist and geneticist.

More recently, a new microarray scanner has significantly boosted speed and efficiency. Morton credits the GeneChip Mapping 10K microarray system, donated by its manufacturer Affymetrix, with speeding the discovery of genetic roots to six rare diseases over the past year. Most recently, the clinic discovered a mutated gene that causes a developmental delay disorder.

While there are roughly 30,000 genes in the human genome, any one of which could have caused the disease, the clinic staff was able to narrow its search to 55 genes on chromosome 6 by scanning 10,000 genetic variations from the patient and two affected cousins and each of their parents. The team finally determined that one gene, SLC1745, was mutated in the developmentally delayed patient. The mutation turned out to be identical to one that causes Salla disease, a rare disorder that has occurred in Finland.

Kevin A. Strauss, a pediatrician at the clinic, says it took less than a month and under $4,000 to make a definitive diagnosis of Salla disease. The patient, a Mennonite girl, had previously undergone four years of standard medical exams, costing nearly $20,000, with no conclusive diagnosis. The microarray scanner also expedited genetic discoveries at the clinic in the areas of Down syndrome, cortical dysplasia, and pretzel syndrome, a previously undescribed disorder characterized by skeletal deformity and brain malformation.

Morton says the clinic sees about 800 patients with about 80 diagnosed genetic disorders, and has cataloged information on several undiagnosed diseases. With access to local genealogies and microarray technology, the clinic for special children is amassing data useful in studying the genetic origin of disease in the broader population, according to Morton.

LineaGen's partners attest to the headway the company has made in accommodating commercial research. Many large medical centers haven't figured out how to manage partnerships with complex technology transfer, says Thomas J. White, chief scientific officer at Celera Diagnostics, a joint venture between Applied Biosystems and Celera Genomics. They have no experience with multi-institutional collaboration.

LineaGen, he says, has expertise in both areas, and can help partners follow ethical guidelines in dealing with the kind of data in the Utah repository. UPDB operates on an opt-in principle, whereby individuals have to give consent for their personal information to be used by the database.

According to White, Celera Diagnostics is studying genetic variations in an inflammatory autoimmune disease related to psoriasis. LineaGen is providing Celera Diagnostics with DNA samples from affected individuals, corresponding clinical data, and controls from UPDB. Celera Diagnostics plans to analyze these samples with the goal of identifying genetic markers for use by Celera Genomics.

Battelle contracted with LineaGen on work to identify diagnostic biomarkers for chronic obstructive pulmonary disease and other respiratory ailments, according to R. Mark Gritz, vice president of health and life sciences at Battelle. The database is a fairly unique resource for looking at genealogy and familial linkages, and the fact that they've tried to link together with health records maintained at the University of Utah is also unique. He adds that Utah has also contributed to another resource tapped by Battelle, the National Lung Health Study, which is funded by the National Institutes of Health.

Gritz cites the work with IBM as another potential plus. It seems like some of what LineaGen and IBM are working on will make our joint research much easier to do, he says. Formatting data from lineage studies is extremely difficult due to the sheer amount of raw data involved and the need to find linkages. You go with a small subset of the database to start with, Gritz says, but initial research generates more questions. You always need new data elements that you never thought you might need in the beginning. I understand that the IBM project, if completed, will allow us to work much faster with less human intervention.

White says Celera is monitoring the work IBM has started. He agrees that a standardized architecture will aid researchers, but he does not see an end to in-house techniques of managing data. He notes that Celera recently devised a proprietary tool called Genetic Risk Score, a means of identifying individuals at elevated risk for myocardial infarction that derives from a study the company published this year on novel genetic markers.

According to Brett J. Davis, marketing manager for information-based medicine at IBM, while the system being developed with UPDB will incorporate standards, it will also allow for customization. We work side by side with UPDB researchers to enhance the model, particularly on the front end where people interact with the data. Davis, however, argues there is an urgent need for standards. We see a convergence between what has traditionally been the health care domain and the research domain, he says. To bridge that gap you need an information infrastructure that is based on standards.

LineaGen's Prescott agrees, pointing to the need to automate data search regimes, which are now generally designed by researchers from scratch on a project-by-project basis. He explains that the technical barriers to developing such an IT architecture will often be easier to surmount than cultural and economic barriers for a system that is envisioned as spanning the health care gamut from researcher to patient.

Every project in a university has some kind of project management, Prescott says. It's basically, I run my lab; you run your lab; let the best man win.' This, he says, won't work with Merck or Amgen. While coordinating efforts to execute a contract is something that many university engineering departments have figured out, Prescott says most medical schools have a long way to go.

Having spent most of his career in academic and institutional research and management positions, Prescott says he is now focused on making a vast research database business-friendly. He notes that the Huntsman Cancer Institute has done a lot of work with corporate partners over the years, exposing him to some of the challenges ahead. And while he is not a GE Six Sigma black-belt-process guy, he claims to be naturally attuned to the niceties of commercial project management: It's just inborn, I think.