Nature has a news story out which highlights the truly frightening dark-side of human genetics research:
Hungary’s Medical Research Council (ETT), which advises the government on health policy, has asked public prosecutors to investigate a genetic-diagnostic company that certified that a member of parliament did not have Roma or Jewish heritage.
The MP in question is a member of the far-right Jobbik party, which won 17% of the votes in the general election of April 2010. He apparently requested the certificate from the firm Nagy Gén Diagnostic and Research, which rents office space at the prestigious Eötvös Loránd University in Budapest. The company produced the document in September 2010, a few weeks before local elections.
The “good news,” if it can be called that, is that the MP in question was scammed by the genetic testing firm that certified him as Jewish/Roma-free. I say this because such a claim is impossible to make and reflects a fallacy in the understanding of categorical genetic variation.
Genetic testing can very reliably distinguish between individuals (e.g. forensic genetics). It can also identify patterns of individual ancestry in increasingly nuanced ways. What it cannot possibly do is definitively rule out ancestry from a given cultural group. Roma and Jewish are cultural categories…they are not defined by the genetic variation they encompass.
Geneticists can look at a sample of Roma and say that a set of genetic traits are at a particular frequency within their sample, potentially serving as a proxy signal for modal tendencies in Roma genetic variation. Geneticists can also identify genetic polymorphisms that, to the extent that other populations have been properly sampled, might be unique to Roma (these ancestry informative markers are the basis for a lot of genetic ancestry testing). Neither approach could definitely rule out Roma ancestry for two important reasons.
The first is that the logical progression in genetic identification is cultural category => genetic variation, not genetic variation => cultural category. A simply example illustrates why this is the case. Every one of us is born into the world with something like 50-200 unique mutations. These mutations may be completely novel at the individual level or repetitions of mutations found elsewhere in the human species (with 3.5 billion base pairs there is a very low likelihood of repeating a mutation, but with 7 billion people on the planet, that likelihood becomes more of a probability-defined certainty). Either way, they represent variation not contained within our parents. Or, likely, their parents. Or their parents. The point being, that child has genetic variation not typical of whatever prior identified genetic signal exists for the Roma or Jewish population in question. Cultural categories form the basis of our identified genetic categories, not the other way around. Expressed more quantitatively, a genetics firm could say their estimate for you ancestry of given group might be 0%, but the confidence interval for that estimate cannot possibly end at 0%.
The second reason the concept fails is even more fundamental. Your genetic ancestry is just a fraction of your total ancestry. This is also easy to conceptualize. I have two older brothers. The three of us have an identical set of ancestors…the same parents, grandparents, great-grandparents, and on and on…We do not, however, have the same genetic ancestry. If we did, we would be pretty close to genetically identical twins (excepting unique individual mutations). We are not genetically identical because, by chance alone, we inherited different parts of each of our parents genomes, who each inherited different parts of their parents genomes. My individual ancestry reflects a probabilistic amalgam of my actual set of ancestors. Shifting the example to the Hungarian MP, he could have Roma or Jewish ancestors whose genes, by chance alone, did not make it to him (accepting that identifying such a genetic signal would even be possible).
The sum of these two points is that there is no such thing as genetic purity. It could not possibly be identified if it did exist, but biology does not create purity, it creates a constantly changing array of variation.
