health

I got thinking this evening about APOE, which includes a very well-known polymorphism of three alleles, where the most ancient (ApoE4) is associated significantly with Alzheimer's Disease risk in European population samples. The association is not significant in all genetic backgrounds, including African American population samples, so it's not necessarily a case where we could predict the phenotype of an ancient genome from observing the allele. But it is one of the most commonly known disease risk polymorphisms, and I hadn't happened to look it up to see what Neandertals and Denisovans are like.

There are two constituent SNP loci -- rs429358 and rs7412. For both these loci, the Denisova genome data include one relevant read, together indicating the ApoE4 allele. The alignment quality of these reads is indicated as poor and I wouldn't take the result to the bank. A third locus, rs4420638 in the nearbyAPOC1 gene is typically linked to the APOE status in living people, and four Denisova reads indicate the allele that is today usually linked to ApoE4. The Neandertal data have no reads at all for the two key SNPs in APOE, and only a single read for the linked SNP in APOC1 is likewise the one usually linked to ApoE4.

None of this is surprising, because ApoE4 is the more ancestral allele. Still, the other common alleles (ApoE2 and ApoE3) are relatively ancient as human polymorphisms go, and could very well have existed in populations contemporary to Neandertals and Denisovans, or in some individuals in those populations. But as it stands, the data suggest that the Denisova genome carried an ApoE4 allele.

An interesting read this morning from Fiona Fox, chief executive of Britain's Science Media Centre: "What If There Were Rules for Science Journalism?"

She proposes a "checklist" for science reporting, which sounds to me a bit like the "Nutrition Facts" that the government puts on a box of cereal.

A checklist would look something like the following. Every story on new research should include the sample size and highlight where it may be too small to draw general conclusions. Any increase in risk should be reported in absolute terms as well as percentages: For example, a "50 percent increase" in risk or a "doubling" of risk could merely mean an increase from 1 in 1,000 to 1.5 or 2 in 1,000. A story about medical research should provide a realistic time frame for the work's translation into a treatment or cure. It should emphasize what stage findings are at: If it is a small study in mice, it is just the beginning; if it's a huge clinical trial involving thousands of people, it is more significant. Stories about shocking findings should include the wider context: The first study to find something unusual is inevitably very preliminary; the 50th study to show the same thing may be justifiably alarming. Articles should mention where the story has come from: a conference lecture, an interview with a scientist, or a study in a peer-reviewed journal, for example.

I think these are good recommendations for health reporting. An awful lot of people have adopted diet recommendations that at best can lower disease risk by a small fraction. Meanwhile, many continue smoking despite much larger and repeatedly demonstrated risks. Science and health reporting have not historically helped people to understand relative risks, and they do a poor job of informing people how scientific conclusions are produced. This lack of transparency has enabled a large niche for "health advisors" who are essentially quacks. People are poorly informed about how to distinguish quack advice from science.

Nevertheless, I think some of Fox's recommendations verge on censorship -- their aim is to stop the public from being misdirected to unreliable findings, but the solutions are all oriented toward stopping the reporting of unreliable findings. I would prefer to see a change in emphasis away from reporting findings and toward reporting process. Scientists trust science without trusting every result, because they understand the process of science. The public will be better informed about scientific results when they see the process in action. A sharp reporter should not only attend to the immediate result of a study but the process underway to test and possibly reject today's findings.

Ann Gibbons reports [1] from the International Congress of Human Genetics, on papers that examine GWAS risk alleles for type 2 diabetes: "Diabetes Genes Decline Out of Africa" (paywall).

At the poster session, Stanford graduate student Erik Corona stood in front of a Google Earth map of the world that he finds surprising. On this map he had plotted the frequency of 12 gene variants known to be associated with type 2 diabetes in 51 populations from Australia to Zaire. It shows “a clear gradient of red to green from west to east, from Africa to Asia,” Corona says (see map). “Something strange is going on with type 2 diabetes.”

This is of course a challenging problem because risk alleles identified in one population may not replicate in other populations. The most well-known example is ApoE4, strongly associated with Alzheimer's Disease in Europeans, but not in Africans. More generally, looking at a set of risk variants that are identified in one population introduces an ascertainment bias that constrains their likely frequencies in other populations. An allele is more likely to yield a statistically significant association with a trait if the allele is not too rare. If we take many alleles associated with a trait, we're likely to see some gradient across populations due to this bias alone.

Hidden ascertainment bias is a problem we run up against quite a lot. It may not apply in this case, depending on where the risk alleles were identified, in particular since many risk alleles for type 2 diabetes appear to be linked to recent positive selection (explaining why I got interested).

References

The 23andMe blog reports on a recent genome-wide association study of type 2 diabetes in South Asian people: "SNPWatch: Genetic Variants Associated with Type 2 Diabetes in South Asians and Europeans". The study was published in August in Nature Genetics, by Kooner and colleagues [1]. As described in the post:

The authors behind this study carried out one of the largest type 2 diabetes studies to date, scanning the genomes of nearly 19,000 people with the disease and 40,000 without it, all of South Asian descent. Their analysis identified six SNPs linked to this condition. When they combined their results with previously published findings in other ethnicities, they found suggestive evidence that five of the six SNPs were also associated with type 2 diabetes in European populations. Similarly, there was some evidence that the majority of the genetic risk factors in Europeans were also linked to disease in South Asians. Only three genetic factors were not shared at all between the two groups.

Type 2 diabetes is presently a very interesting topic from an evolutionary viewpoint, and we're beginning to think about it very seriously now. Whenever I see a study like this, I quickly look at the Neandertal and Denisovan genomes to see if any interesting patterns emerge. Sharing GWAS SNP alleles is not necessarily very interesting, because the GWAS risk alleles are mostly not causative themselves; each may be linked to some causative allele that remains to be discovered. The linkage is a function of the evolutionary history of that chromosome region, and many of the key historical events that affect linkage happened within the last 10,000 years. So we really shouldn't expect GWAS alleles to be predictive of phenotypes in Neandertals or Denisovans.

Still, these alleles are associated with disease in living people, and their genotypes in ancient humans may illuminate cases where the evolutionary history links the population across the gene networks that influence disease. A closer examination of the genealogy around these loci will be more informative, but as a first look I often just genotype the archaic genomes for SNPs in a study. The six SNPs reported here include two cases where the archaic genomes have the derived risk alleles, one of them present in Neandertals but not the Denisova genome. Again, that doesn't tell us anything about the phenotype of the ancient people, but worth a closer look to see if one or both of these is an introgressive allele.

We have here the GWAS Catalog genotypes for all the archaic genomes. Not much actionable information but there are some interesting phenotypes in there. I'll share some more of those later this week.

References