In an era of genetics research where huge cohorts seem to be the name of the game, two new papers highlight the power of small. Reported online March 10 in SciencExpress and the New England Journal of Medicine, the independent studies used whole-genome sequencing within a single family to identify disease genes for rare heritable conditions. The procedure is expensive and likely more challenging for disorders with complex phenotypes. However, some scientists say the recent advances have set a new benchmark for genetics research and diagnosis. One of the research teams plans to apply whole-genome analysis toward studies of neurodegenerative disease.
In the SciencExpress paper, researchers led by Leroy Hood and David Galas, Institute for Systems Biology, Seattle, analyzed whole-genome sequences of a family of four—two children with the recessive disorders Miller syndrome (characterized by abnormalities in the head, face, and limbs) and primary ciliary dyskinesia (which impairs mucus clearance from the lungs), and their unaffected parents. Jared Roach, Gustavo Glusman, Arian Smit, and Chad Huff were joint first authors on the research, which included collaborators at Complete Genomics, Inc. in Mountain View, California.
Family genome analysis affords unprecedented accuracy—99.999 percent in this case, the authors report. By reading all six billion nucleotides of both parents and children, the researchers were able to trace the origins of each chunk of DNA in the kids. “This means that you can just look at the sequence, and whenever you see something that doesn’t make sense from a Mendelian point of view, chances are almost 100 percent that it’s sequencing error,” Galas said. The ability to weed out these glitches boosted their accuracy about 10-fold above that of typical sequencing efforts, he noted. This allowed the researchers to narrow the field of gene candidates for Miller syndrome to just four, whereas the list would have had 34 if the study had not included the parents’ genomes. In addition, the analysis turned up a gene previously identified as causing primary ciliary dyskinesia (Olbrich et al., 2002), providing further evidence that pathogenic mutations of rare disease can be identified by complete genome analysis of small families.
Linda Avey, who co-founded the Mountain View-based personal genomics company 23andMe, said the study represents “tremendous technological progress.” It is “something we’ve been waiting impatiently for,” Avey, who left 23andMe last fall to start a foundation focused on cognitive health, told ARF. “It’s hugely important and will change the field of genetics research.”
The advance comes with a fearsome price tag. The sequencing cost about $25,000 per genome in this analysis in Science, and roughly $50,000 for the genome sequenced in the NEJM study. That DNA came from first author James Lupski, a geneticist at Baylor College of Medicine in Houston, Texas. Born to healthy parents, Lupski and three of his siblings have the recessive form of Charcot-Marie-Tooth disease, a nerve disorder marked by loss of muscle tissue and touch sensation. Richard Gibbs, another Baylor geneticist, led the team that sequenced Lupski’s genome. The scientists focused on mutations in 40 genes with known connection to neurologic diseases, and genotyped these variants in affected family members. The analysis pinned Lupski’s disease to a gene that seems important for nerve transmission in mouse studies.
The authors note that DNA read lengths doubled (from 25 to 50 base pairs) and sequence yield went up threefold during the six-month course of the study, reflecting the “current rapid progress of sequencing technology.” Galas estimates that by the end of the year, sequencing costs will drop to $5,000 a genome, and that in a few years, “it could be a few hundred dollars, which is around what it costs for genotyping now.” George Church of Harvard Medical School, Boston, concurs, noting that sequencing costs have plummeted 40,000-fold over the past five years (see full comment below).
However, even if technology continues to drive down sequencing costs, whole-genome analysis may prove harder for complex disorders such as Alzheimer disease. It's one thing to correlate genetic variants—whether collected via SNP genotyping or full sequencing—to clear-cut disease phenotypes, such as those studied here, but quite another for complex diseases, like Alzheimer's, where the diagnosis and phenotype are less straightforward and where the cause is a combination of genes and environment, Avey wrote in an e-mail to ARF (see full comment below and blog post on the importance of phenotypic data). Furthermore, the parents of a majority of people with AD are deceased, so family genome sequencing in these cases would look different: the patient would be the parent, and their healthy adult offspring would presumably receive some sort of risk prediction through the genome analysis. However, some patients with early onset AD have parents who are still alive.
Acknowledging the added challenge that is sure to come with genome studies of AD and other conditions such as schizophrenia, for example, Galas said he still believes the current analysis provides proof of principle that whole-genome sequencing is “applicable to any disease no matter how complex.” Moving forward, his team plans to test the approach for studies of neurodegenerative disease. Initially, they will focus on Huntington’s, which is simpler than Alzheimer’s or Parkinson’s in that it has one gene (huntingtin) with a major effect and other modifier genes that change symptoms and severity, he told ARF. “What we're hoping to do with that is take one baby step toward more complex genetics and try to find modifier genes,” he said.—Esther Landhuis