March 31, 2011
Rare Genetic Variants Make Biggest Health Impact
The rare genetic variants and not the common variants contribute the most to disease risks.
DURHAM, N.C. – New genomic analyses suggest that the most common genetic variants in the human genome aren't the ones most likely causing disease. Rare genetic variants, the type found most often in functional areas of human DNA, are more often linked to disease, genetic experts at Duke University Medical Center report.
We all carry at least hundreds of rare genetic variants. So one has to read "rare" to mean that each rare variant is not covered by large numbers of people even though the total number of rare variants is very high.
These results make sense because any genetic variant that makes a big negative impact on health will usually get selected out of a population before the variant spreads to large numbers of progeny.
The study was published in the American Journal of Human Genetics on March 31.
"The more common a variant is, the less likely it is to be found in a functional region of the genome," said senior author David Goldstein, Ph.D., director of the Duke Center for Human Genome Variation. "Scientists have reported this observation before, but this study is the most comprehensive effort to date using annotations of the functional regions of the human genome and fully sequenced genomes."
Goldstein said that "the magnitude of the effect is dramatic and is consistent across all frequencies of variants we looked at." He also said he was surprised by the notable consistency of the finding. "It's not just that the most rare variants are different from the most common, it's that at every increase in frequency, a variant is less and less likely to be found in a functional region of the DNA," Goldstein said. "This analysis is consistent with what appears to be a growing consensus that common variants are less important in common diseases than many had originally thought."
This makes the discovery of the genetic variants that impact disease risks much harder to do. The number of disease-risk influences is much larger. Each variant has very few people carrying it. So there's less of a mutual benefit when each harmful variant is found.
This result is an argument for allowing the general population to spend their own money to get themselves genetically tested and sequenced. They can then submit their own genetic test results and disease history to medical genetics research groups for study. To discover all the genetic variants that influence disease risks will require most of the population to get themselves genetically tested and then to share their genetic test results with medical genetics researchers.
Rare variants, not common ones, contribute most to disease.
Well, big f-ing deal and ring all the villages' bells. It took all this time and all my tax dollars for groups of grant-seeking, prestige seeking geneticists who evidently have no knowledge of natural selection to conclude this? What fields of study do geneticists actually "study," might I ask?
They ought to be replaced with people with at least a basic knowledge of evolution, for God's sakes.
In the early days geneticists did find common variants with major disease risk. E.g., APOE4 for Alzheimer's Disease and BRCA1 for cancer. Hence the prediction that with more extensive testing for genetic markers then common variants would be discovered for many more diseases.
The main model of human genetics has been that most human variation is due to genetic drift. In this model there has been little selection over the last 100,000 years. This model acknowledges that there has been selection in specific cases such as skin color, lactose tolerance, and disease resistance but views such cases as exceptions to the general rule. This was the genetics model driving the genetics projects whose results have been reported over the last few years. The more recent model of accelerating human evolution with increases in population size was published after those projects were initiated and is not yet the consensus model for human evolution.
The modern environment differs greatly when compared to that of 100,000 years ago. On average people live far longer, eat completely different diets, are exposed to far more pathogens due to high density cities and global transportation, are much less physically active, and have new patterns of differential reproductive fitness (teenage welfare moms). Modern humans have not had time to genetically adapt. (More precisely, the adaptations are often at an early stage in which there are many secondary, harmful affects.) There has not been sufficient time to filter out common variants that could be causing modern health problems associated with, e.g., obesity. Hence the assumption that more extensive searches would uncover common DNA variants as major risk factors for important diseases wasn't ruled out by evolutionary theory. Note that this is true in both the drift model and the accelerating evolution model.
So why were the predictions wrong?
First, I'd ask if they really are wrong. In the genetics literature "rare" usually means that the variant occurs with a frequency of under 5%. This cutoff is due to the size of the typical study population and the need to meet the "statistically significant threshold" P.05 (which is really much too weak). Once complete genetic data is available from millions of people then "rare" might mean one in a thousand or one in a million. With that shift in meaning scientists may yet find that a variant that occurs in millions of people worldwide is a major contributor to some common disease (each such variant would only account for a tiny fraction of the world's total disease variance but for those with the variant it would be very important). I believe this is true for some forms of Type I diabetes.
Second, most important traits are complex. I.e., they depend upon the proper functioning of hundreds to thousands of distinct components. A simple trait would be skin color whose variance in humans largely depends on a few variants. In contrast a complex trait such as intelligence will depend upon the interaction of thousands of genes and gene products. By necessity complex systems must be robust in the face of significant genetic variation. (Otherwise sexual reproduction in large populations of genetically diverse individuals would commonly result in unfit children.) Hence most common genetic variants have only modest affect on complex traits and complex traits depend on hundreds to thousands of variants.
So why didn't the scientists predict that it is unlikely that common variants explained a significant portion of the variance in major diseases? If I recall some scientists did make that prediction but didn't have an alternative research plan that aligned with the goals of the NIH. Also it is only in the last decade or so that biologists are beginning to get a handle on how robust regulatory genetic control networks arise and evolve in eukaryotes. When you don't know how the underlying molecular biology functions you can't know what will or won't be found. Some times the best strategy is to just do the experiment and see what you discover. (E.g., the experiment that led to iPSCs.)
Interesting. Thanks for the comments, Fly.
Actually, and I have to say it initially surprised me, erica is largely correct. Lots of prominent geneticists do in fact know nothing about natural selection. For example, Kari Stefansson, head of Decode in iceland, was surprised to find that rare variants played a strong role in schizophrenia, which implied that those variants stayed rare because of selection. He thought that "the brain was a luxury organ when it came to natural selection. " Truly, a maroon.
The whole common disease/common variant notion never made any sense - except for diseases of old age, where selection is weak. Which is why positive results about common variants like APOE4 and Alzheimers are found, similarly with macular degeneration and late-onset glaucoma.