On July 28-29, 2014, the National Human Genome Research Institute (NHGRI) convened geneticists, biologists, clinicians, bioinformatics specialists and others to provide expert advice on the future of its flagship Genome Sequencing Program (GSP).
The two-day meeting, titled, Future Opportunities for Genome Sequencing and Beyond: A Planning workshop for the National Human Genome Research Institute, was the third formal review of the GSP, the previous discussions taking place in 2005 and 2009. After the last review, new core programs were added, including efforts to find causes of rare diseases and move genome sequence information into the clinic. In addition to the long-established Large-Scale Sequencing and Analysis Center (LSAC) program, the GSP now includes the Centers for Mendelian Genomics program, the Clinical Sequencing Exploratory Research (CSER) program and the Genome Sequencing Informatics Tools (GS-IT) program.
NHGRI should continue its role as a "genomics trailblazer," said NHGRI Director Eric Green, M.D., Ph.D. This means supporting research into new DNA sequencing methods and data analysis techniques and solving difficult problems in genome sequencing, he said.
The success of NHGRI's efforts in this area will lie in its ability to partner with other institutes and organizations, he said. As genomics moves increasingly into the clinic, new strategies and partnerships will be needed to expand its presence and explore its usefulness and cost effectiveness. Participants and speakers echoed this theme throughout the workshop.
The morning of day one focused on the state of the science. Michael Boehnke, Ph.D., University of Michigan, Ann Arbor, reviewed the status and opportunities surrounding the discovery of variants conferring risk for common disease. He said that one goal for the next five years should be to advance the understanding of the genetic bases of human disease and use that knowledge to improve human health.
Choices for investing resources have to be made, he said, noting that studies should span a spectrum of disease classifications. Variant discovery is only in its infancy, and large studies will be needed to identify disease-associated variants.
Another discussion centered on Mendelian diseases, which are rare and typically due to alterations in a single gene. The Centers for Mendelian Genomics have made discoveries of novel genetic bases or novel clinical features of more than 280 Mendelian diseases since their launch two-and-a-half years ago. The three Mendelian centers are at Yale University, the University of Washington and a joint center at Johns Hopkins University and Baylor College of Medicine.
Studying Mendelian genes and variation contributes to the understanding of complex diseases, said speaker Roderick McInnes, M.D., Ph.D., Lady Davis Institute for Medical Research in Montreal, Canada. "Identifying these genes enhances the understanding of human disease and biology. Because most Mendelian conditions are extremely rare and found in only a handful of families, Mendelian diseases and conditions are the epitome of personalized medicine."
NHGRI supports several genomic medicine programs to explore the best ways to move genome sequencing information into the clinic and to understand how that information may be used most effectively. Understanding and interpreting genomic variants poses one of the greatest challenges to implementing genomic medicine, said Dan Roden, M.D., Vanderbilt University, Nashville, Tennessee. "We don't know the function of many of the variants we identify. If we get this wrong, there is the potential for a lot of people to have information about rare variants without understanding the implications."
In addition, there needs to be more research on the outcomes and cost effectiveness of using of clinical genome sequencing. This research should help determine the value of whole genome sequencing (examining all of the genome, not just the protein coding area), how to identify actionable genetic variants (which may be useful in patient care) and what to do with incidental findings (unintended DNA sequencing results that might raise the risk of other conditions or diseases).
"If NHGRI doesn't take a coordinated action, the promise of genomic medicine will be delayed," Dr. Roden said. "We need large studies to work out the realities of implementation. NHGRI needs to continue to play a role in that."
At the end of the first day, attendees spent two hours in breakout sessions, discussing a range of topics, from clinical genome sequencing to comparative genomics.
On the workshop's second day, speakers discussed ways to implement their ideas based on the previous evening's meetings.
In one session, Eric Boerwinkle, Ph.D., the University of Texas Health Science Center in Houston, Texas, discussed the need to better understand the "architecture" of health and disease. He stressed the need for NHGRI to support basic scientists by creating, for example, new tools for large scale genome sequencing, large scale functional characterization of variants, new computational models and research on non-coding variants.
As the use of "clinical and personal DNA sequencing increases, the basic research we do must drive discovery first. This will improve disease diagnostics and translation to therapies," he said. "At the same time, we have to develop an infrastructure so clinical data can drive novel discovery."
Though the use of clinical genome sequencing is rapidly increasing, the vast majority of research into effectiveness and implementation is occurring in academic medical centers. Sharon Plon, M.D., Baylor College of Medicine in Houston, Texas, discussed the need for randomized trials of large-scale sequencing in a variety of medical settings. "Studies need to reflect populations in different health care and clinical settings across the nation," she said. "NHGRI needs to partner with other institutes to pick the best clinical settings and to design trials that show the efficacy of using genome sequencing." It should also fund research to improve methods for implementing genomic sequencing into routine medical practice such as decision support tools, she said.
One issue that came up several times was the influence of the marketplace in driving the use of genomics in clinical medicine, and the role of NHGRI. Many thought that NHGRI should help set the tone by continuing to invest in clinical genomics programs. James Evans, M.D., University of North Carolina, Chapel Hill, said that NHGRI's Clinical Sequencing Exploratory Research program and other clinical genomics programs "have shown that using genomics in the clinic is possible and perhaps even feasible. It's early in the game and outcomes and studies need to be done to set standards and find evidence of what works."
Workshop organizer and GSP program director Adam Felsenfeld, Ph.D., acknowledged the many challenges ahead for NHGRI and the field of genomics. In addition to new opportunities in clinical genomics, particularly in oncology, there are many areas - and questions - to be addressed, he said. He pointed to the need for greater discovery of genetic variation associated with disease, a better understanding of functional genomics and how the genome works, and improved methods of genome sequencing and assembly. He echoed many workshop participants in acknowledging the need for an improved reference genome.
Comparative genomics continues to be a powerful tool for discovery and interpretation, and needs to be supported, Dr. Felsenfeld said. There needs to be better ways to quickly determine if variants found in research or in clinical genome sequencing are useful in patient care, especially for common diseases. As more data becomes available in and outside of NIH-supported programs, an informatics infrastructure will be necessary to make such information widely accessible.
"It's a very exciting time and it's difficult in part because there are so many opportunities," said Dr. Evans. "There are many needs, and many appealing directions that we can go in."
Last Updated: September 15, 2014