Published: September 29, 1997
Bioinformatics looms large
ROCKVILLE, Md. -- Bioinformatics is expected to play an important role in the research efforts of the newly formed North American Rheumatoid Arthritis Consortium (NARAC) here, a coalition of medical research centers and other arthritis-related institutions that has the mission of finding genes that determine susceptibility to the disease. Consortium members include the National Institute of Arthritis and Musculoskeletal and Skin Diseases, the National Institute of Allergy and Infectious Diseases, the Arthritis Foundation, and a dozen research centers nationwide.
The group has set ambitious goals for the next five years. Principal investigator Peter Gregersen of North Shore University Hospital, Manhasset, NY, told Biolnform that the scale of what they hope to accomplish will require significant bioinformatics support. "This project involves, first of all, collecting a large number of families -- 1,000 -- with rheumatoid arthritis, in which there are two siblings affected," he explained. "Then we are planning to do a genome-wide screen for genetic regions that are involved in disease susceptibility. That project will involve probably close to a million genotypes being performed. So obviously there will be a lot of information being dealt with and analyzed for allele-sharing among these sib pairs. We'll be looking for regions that are shared more frequently than you would expect by chance."
Gregersen estimated that 5-10 percent of NARAC's $7.8 million budget will go toward bioinformatics, with some of that amount devoted to development of bioinformatic techniques by consortium members.
Most of the bioinformatics work will be performed or coordinated by a research group headed by Christopher Amos at MD Anderson Cancer Center in Houston. "The immediate roles for bioinformatics occur at different stages," he elaborated. "Obviously there are informatics needs to translate results from automated genotyping machinery into allelotyping results. Our center at MD Anderson will then take the allelotyping findings and use these for genetic linkage and association studies. For the linkage analyses, we assess the genetic similarity among related individuals for each particular chromosomal region and then search for chromosomal regions that display strong similarity among affected individuals within families. This provides a tool for identifying genetic regions that are likely to be involved in causing rheumatoid arthritis."
"Association studies will then be conducted in regions that show excess similarity among affected individuals," Amos continued. "For these analyses, we will evaluate evidence that any particular alleles at the genotyped loci are more common among affected individuals than among controls. The primary method that we are using compares alleles transmitted to affected children to alleles in the parents that were not transmitted to the affected children. The analytic needs for the grant are substantial because of the large volume of markers to be processed."
NARAC's funding doesn't currently provide support for technology development, Amos noted, but he added that as the research progresses, changes could be necessary. "We may be required to develop new analytic strategies to accommodate particular problems that exist in the data," he said. "Our group has already developed some analytical tools for association analyses and these may need to be extended. Database management of the clinical data will be performed by Dr. Gregersen's group at North Shore Hospital, but will not require the development of new technologies. The database management support does require that database management tools be created from standard software, such as Microsoft Access."
New bioinformatics techniques will be applied by other members of the consortium, including Gregersen and Michael Seldin at the University of California, Davis, Amos commented. "For bioinformatics associated with genotyping, both Dr. Seldin's and Dr. Gregersen's laboratories are involved in the application of new genotyping methods," he explained. "Limited support for genotyping-related informatics issues may be provided through collaboration with biotechnical companies. Dr. Seldin currently has some additional grant support related to informatics for genotyping from the human genome initiative."
The NARAC bioinformatics work will be performed with existing equipment, Amos said. Seldin's group will use ABI genotyping machines. Data entry and database management software will be developed using Microsoft Access. Several programs will be used for statistical analyses: GeneHunter, created by Leonid Kruglyak and Eric Lander; SIBPAL, created by the Statistical Analysis for Genetic Epidemiology group headed by Robert Elston; MENDEL, created by Ken Lange, Michael Boehnke, and Dan Weeks; and ACT, developed by Amos's group.
David M. Lawrence