Some further reaction and expanded comments from geneticists and isotope specialists who examined the UK Border Agency documents--a stakeholders letter ( stakeholder+letter.11.9.09.doc) and one titled "Nationality-Swapping ( nationality-swapping-DNA-testing.pdf)--describing the Human Provenance Pilot Program detailed in an earlier ScienceInsider story.
Alec Jeffreys, University of Leicester: Nationality is not genetic, and by the UK Border Agency's logic, huge numbers of people in the US and UK would fail a DNA-based nationality test (there is no defined US ethnic population unless you use indigenous North Americans in which case virtually everyone fails, and much the same is true for the UK). The Border Agency is clearly making huge and unwarranted assumptions about population structure in Africa; the extensive research needed to determine population structure and the ability of DNA to pinpoint ethnic origin in this region simply has not been done. Even if it did work (which I doubt), assigning a person to a population does not establish nationality - people move! The whole proposal is naive and scientifically flawed.
Chris Tyler-Smith, The Wellcome Trust Sanger Institute: I think there is a fundamental limitation in that DNA can (at best) provide evidence about deep geographical ancestry, and not about nationality. mtDNA and Y markers would give only a partial view of geographical ancestry. So I would have reservations about the use of DNA tests for this purpose.
David Balding, Imperial College London: Genes don’t respect national borders as many legitimate citizens are migrants or direct descendants of migrants, and many national borders split ethnic groups.
There’s no reference to any research to underpin all of this, and no details about the type of genetic testing. Unfortunately there are a lot of charlatans in this sort of business, including some in top universities. Although the case managers are told to weigh up all the evidence, and that the DNA and isotope results may be wrong, how can they have any possibility of assessing the reliability of this complex scientific evidence – even the experts are not very good at this.
Christopher Phillips, University of Santiago de Compostela: I had been asked earlier this year by colleagues in the UKFSS about the prospects of differentiating Somali ancestries from other populations in E[ast] Africa, however, I am sceptical about the precision possible beyond a simple five global group differentiation from limited typing of Y-chromosome/mtDNA/small-scale multiplexes of autosomal SNPs. Clearly there is a serious risk of falling into the trap of over-interpretation of population variation data that has limited scope. My suggestion this spring was to perform whole genome scans to isolate informative markers and begin to build these into sets of SNPs that could then be assessed with comprehensive reference populations. However, this does not amount to consultation on the correct way to develop and test a custom ancestry analysis system. I also doubt that my suggested approach to validating the system will be pursued, since a large number of samples would be required both within the relatively large region of Somalia and from surrounding populations such as those of Ethiopia, Sudan and Eritrea. Therefore a good deal of time, money and patience would be needed to find the best markers for the purpose and then test their efficacy.
Mark Thomas, University College London: This is horrifying. mtDNA will never have the resolution to specify a country of origin. Many DNA ancestry testing companies have sprung up over the last 10 years, often based on mtDNA, but what they are selling is little better than genetic astrology. Dense genomic SNP data does have some resolution…but not at a very local scale, and with considerable errors. And obviously such tests could only apply to people whose ancestors were from broadly the same region. Also, some ethnic groups may look genetically very different from neighbouring peoples, and as with the isotopes, baseline data from different regions is required first.
Jane Evans, NERC Isotope Geosciences Laboratory: I can’t imagine how you use [isotope evidence] to define nationality....It worries me as a scientist that actual peoples’ lives are being influenced based on these methods.
James Ehleringer, University of Utah: In 2008 we published a paper establishing a relationship between the hydrogen and oxygen isotope of human hair and geographic patterns…To the best of my knowledge, we are the only laboratory that has taken the hair isotope patterns and explicitly placed them onto a GIS display (i.e., map). As you see from that map, there are not unique relationships in the USA, only banding patterns. I am unaware of any substantive regional and/or global pattern analyses other than those out of our lab. Thus, I find it surprising that scientists in the UK have not only acquired the data to develop global maps but also that this technique has been refined to the point that it is a viable, reliable screening tool with the specificity implicit in the document. There are not unique global distribution patterns for hydrogen and oxygen isotopes of water. Instead there are repeatable patterns across continents in a general sense. We would expect hair hydrogen and oxygen isotopes to exhibit the same pattern. Thus, a unique specificity that would be appropriate for the proposed application seems unlikely. Now the additional analyses of heavy isotopes in hair may further constrain the possible regions of origin, but I do not believe that our science has reached the point where we can be specific at the country or within country scale.
Now to the bigger challenge…Hydrogen and oxygen isotopes of hair reveal recent regional geographic patterns. Assuming that hair grows at about 1 cm per month, a 10-cm hair length might record the last 10 months of that person's travel…this 10-month period is not necessarily synonymous with where that person may have originated from.
Tamsin O'Connell ,McDonald Institute for Archaeological Research, University of Cambridge: I am very surprised that the UK Border Agency appears to have not read the isotope literature. I can’t think of any active [isotope analyst] who would think this is a good idea.
Jessica Pearson, University of Liverpool: It’s incredibly vague about what they’re going to do and their methodology…It’s doesn’t really add up. They’ve put 2 and 2 together to get 3 ½. I’d hate to see asylum decisions made with this. It’s peoples’ lives we’re dealing with.