The National Institutes of Health celebrated the 10th anniversary of the publication of the sequence of the human genome and the subsequent remarkable burst of scientific and clinical discovery at a symposium Feb. 11. Fittingly, the National Human Genome Research Institute took the opportunity to unveil its long-awaited blueprint for genomics research, published in Nature on Feb. 10, and online.
The blueprint, called “Charting a course for genomic medicine from base pairs to bedside,” provides a glimpse into the next decade of genomics research. This new strategic vision has a considerably more clinical slant than its 2003 predecessor, reflecting the growing confluence of genomics research, clinical research and clinical care.
The progression of genomic science is broken into five overlapping phases: 1) understanding the structure of genomes, 2) understanding the biology of genomes, 3) understanding the biology of disease, 4) advancing the science of medicine, and 5) improving the effectiveness of health care.
These phases are predicated on the increasing ability to develop diagnostics and treatments based on fundamental biological insights driven by genomic sciences. Currently, most genomics research is in phases 1-3, but over the next 10 years the blueprint predicts a gradual shift to phases 2-4, and beyond 10 years to phases 2-5. The blueprint portrays non-linear advances, with ongoing clinical advances coming to health care.
General imperatives are also identified, including making genomics-based diagnostics routine, defining genetic components of disease, comprehensively characterizing cancer genomes, developing practical systems for clinical genomics informatics, and defining the human microbiome's role in health and disease.
The document argues that much remains to be learned about the fundamental biology of genomes, which can best be revealed by comprehensive analysis of genomes from many individuals and a diversity of other organisms. New sequencing technologies have turbocharged these efforts, but understanding genomes' fundamental biology requires new approaches for high throughput measurement of the epigenome, proteome, and RNA molecules of a wide variety of cell types in a wide variety of states. This in turn requires developing large numbers of “affinity reagents,” antibodies for example, and new techniques to facilitate studies on small numbers of cells of small samples.
In the area of biology of disease, similar scientific resources will need to be brought to bear. A pressing and very feasible near-term priority is the elucidation of the spectrum of primary sequence variations underpinning the remaining uncharacterized Mendelian conditions, of which there are several thousand.
Somatic variation in cancer has been shown to be immensely complicated, and working out these complexities will require time and innovation. Already, genome-wide association studies have provided a wealth of new information about common, complex conditions such as coronary artery disease and type 2 diabetes. However, much remains to be learned about variations aside from common single nucleotide polymorphisms associated with disease. Sequencing costs are no longer the prohibitive bottleneck that they were five years ago. That bottleneck is now the collection and preparation of accurately phenotyped samples, as well as data analysis for large numbers of individuals.
Another looming challenge—made easier by comprehensive catalogs of normal variation—will be determining the causality of variations detected in association studies of affected and control subjects, particularly in complex disease. The biggest hurdle may be the massive amount of work needed to determine genes' function and related pathways in normal as well as disease states. This can be facilitated by developing libraries of chemical probes (amounting to smart bombs) that interact with proteins and possibly functional RNA molecules in highly specific ways to disrupt function.
The new blueprint for the future of genomics articulates a clear call for evidence of clinical utility for new genomic health applications. Novel methods for clinical trial design are particularly needed to circumvent the prohibitive expense of gathering large numbers of individuals with similar genotypic profiles for traditional trials. As I have written recently, new informatics applications will be required to effectively analyze the volumes of data from the resulting studies. Electronic health records will need to become genome-enabled to effectively store, analyze and help clinicians make use of genomic data. Clinicians will also need to become more genome-enabled, likely through electronic clinical decision support systems.
The blueprint for genomics research highlights the increasing importance of addressing issues at the intersection of genomics and society. Widespread appropriate use of genomic advances will require clinicians, patients and policymakers to understand the potential benefits and harms that come with all technological advances. Likewise, public policymakers require sound science that addresses the legal, ethical, and health service delivery issues that are increasingly raised by new genomic applications. In the end, these issues will be as important as the basic science in realizing the benefits of genomic advances in the decades to come.