Exaggerations and errors in the promotion of genetic ancestry testing

Early_diversification

 

One thing we have done in Genomes Unzipped is to report on what is on the market for consumers interested in getting information about their genetic data. While we have found generally positive things to say about this market, there are also many exaggerated claims especially when it comes to making inferences about an individual’s ancestors from direct-to-consumer genetics companies. An example came up last summer with a BBC radio 4 interview of Alistair Moffat of Britain’s DNA.  This post will discuss the scientific basis of some of the claims made in the interview.

But first of all, what is my motivation to write this post? After all, there are quite a few genetic ancestry companies like Britain’s DNA, making similar claims. Why specifically discuss this BBC radio 4 interview? The main reason is that listening to this radio interview prompted my UCL colleagues David Balding and Mark Thomas to ask questions to the Britain’s DNA scientific team; the questions have not been satisfactorily answered. Instead, a threat of legal action was issued by solicitors for Mr Moffat. Any type of legal threat is an ominous sign for an academic debate. This motivated me to point out some of the incorrect, or at the very least exaggerated, statements made in this interview. Importantly, while I received comments from several people for this post, the opinion presented here is entirely mine and does not involve any of my colleagues at Genomes Unzipped.

To give some background, Alistair Moffat (AM), who is also the rector of the University of St Andrews, is involved in a trio of related businesses:  Britain’s DNA, Scotland’s DNA and Ireland’s DNA. These companies will type Y and mtDNA markers for their customers and provide them with an interpretation of their ancestry based on these results. The cost of the service is £170 for either Y or mtDNA testing; there is a reduction if you take both.  This price is somewhat comparable (actually a bit higher) than other companies generating similar data (for example familytreedna.co.uk). A claim made in the BBC radio 4 interview is that this cost is massively subsidized but given that the test is not particularly cheap it is not clear what this statement really means. I will get back to this issue at the end of this post.

A bit of clarification on chromosome Y and mtDNA: these data represent only a small portion of the human genome and only provide information about the male (fathers of fathers of fathers…) and female (mother of mothers of mothers…) lineages. As an illustration, going back 12 generations (so 300 years approximately) we each have around 4,000 ancestors. mtDNA and chromosome Y DNA only provide information about 2 of them. So these markers provide a very limited window into our ancient ancestry. As a comparison, the genome-wide data provided by companies like 23andMe is much more exhaustive. This is however not a central point of this post. Britain’s DNA is very explicit about the data they provide, but it is important for readers of this post to be aware of this limitation.

Before discussing this radio interview, I will make two last comments. Firstly, this post is not a review of the commercial product, which I have neither purchased nor tested. It is plausible that AM genuinely misunderstood scientific facts and that his team is more careful with the interpretation of the data. A proper review of the product would be the topic of another post. Secondly, what makes this interview a disaster is also the BBC host, Jim Naughtie, who asked no sceptical, probing or challenging question in the face of claims made by AM that many scientifically minded listeners might recognise as dubious (see below). This is very disappointing for such an influential radio programme with a broad audience: Naughtie seems to have taken the view that because it was science he couldn’t possibly be expected to probe, but if nothing else a broadcaster can always ask “what is the evidence?”

Now after this long introduction, let us go through some of the statements made by AM in the BBC radio 4 interview. An ongoing theme of the interview is the link with biblical facts and how some of his findings are consistent with a biblical interpretation. So AM starts with:

“It has been posited by scientists for about 20 years that Adam and Eve really existed. They  may never have met, but they really existed”.   

This sort of statement starts the interview on the wrong foot. Surely, basic population genetics provides some basis for concepts interpretable as “Adam” and “Eve” if we can assume that AM is referring  to the most recent common ancestor at the mtDNA and at the Y chromosome. The existence of such a common ancestor is an obvious fact for population geneticists, so it is pointless to mention any time such as 20 years for scientists  “positing” what is simply a matter of definition. And while one can define something that looks like the concept of Adam and Eve, there is really nothing particularly special with these individuals. The human genome must have a large number of “ancestors” at various genomic positions. mtDNA and Y chromosome represent only a small portion of the human genome, and being the ancestor at either of these locus cannot be interpreted as supporting any biblical interpretation.

“What happened was a kind of genetic bottleneck – around 70 thousand BC an Indonesian  volcano called mount Toba blew itself to smithereens and it was almost a species extinction  event and so all the other lineages apart from these two, apart from Adam and Eve were  destroyed”.   

While the Toba eruption may not be controversial, its effect on human population size and  on patterns of genetic variation certainly is. More importantly, no matter how devastating  its effects were, there is no reason to believe that this event caused all lineages apart from those of “Adam” and “Eve” to be destroyed, and indeed it is highly unlikely that this was the  case.

After this general introduction, AM moves on to describe how this relates to the Britain’s DNA results:

“And what happened with the Britain’s DNA project is that about 4 or 5 weeks ago we  discovered a remarkable individual, a Mr Ian Kinnaird from Caithness and he has Eve’s DNA  – he’s only two removed from Eve … he carries a marker called L1b which is only two  mutations different from what Eve’s marker must have been … he’s Eve’s grandson …”.

Firstly, while we discuss above that one can put some meaning to the concept of “Eve” in population genetics, this is far removed from the biblical interpretation implied here. Secondly, while there has been some prior attempts to estimate Eve’s mitochondrial DNA sequence, there is quite some uncertainty about these sequence. Lastly, being two mutations away from Eve is highly implausible: we are all separated from Eve by the same amount of time and approximately the same number of generation. Given the high mutation rate of mtDNA, I strongly suspect that there is something technically wrong with this computation (but of course I have not seen these data).

“..but we found Sheban DNA –a marker called HV– which we didn’t expect to find, and I say we’ve got nine people who…”

There is no solid scientific justification for claiming that the HV marker (which is usually referred to as haplogroup rather than “marker”) corresponds to “Sheban DNA”. We don’t know what the haplogroup of the Queen of Sheba was, this haplogroup is dispersed in the Middle East and other regions and it is not unsurprising to find individuals with this haplogroup in the UK.

In fact this is an opportunity to address a key issue with that set of interpretations, which has been labelled as “interpretative phylogeography”. The underlying rationale that is behind these statements is that a haplogroup must have originated from the point of the world where it is most commonly found today. While this may be true in some cases, there are many reasons why this can be incorrect. In particular, random drift of alleles or population migration can easily make allele frequencies vary extensively, which will cloud that signal. Moreover, our knowledge of the world’s genetics is limited and some population have been studied much more than others, hence are more likely to be assigned a “founding role”.

“Men with the name Cohen, 97%, a huge number, share the same marker…”.  

A UCL colleague (Mark Thomas) was the first author on the original paper that estimated a common ancestry date for the majority of Cohanim (Thomas MG, Skorecki K, et al (1998) Origins of Old  Testament priests. Nature 394:138‐140). While the data support some shared haplogroup among Cohanim individuals, the numbers are nowhere near 97%. This haplogroup is also found in many Jewish and non-Jewish individuals who do not have that last name. Hence such results are difficult to interpret.

“…they call themselves the sons of Aaron, the first high priest, and so as I say, the Bible, through the Britain’s DNA project  and other research is really beginning to come alive…”.

This statement recapitulates some of the extrapolations discussed above. While putting genetic data in that biblical context is attractive, there are few scientific facts to support that interpretation. And in any case, Britain’s DNA has not contributed to these research questions, unlike what this statement seems to imply.

“we found that 33% of men are closely associated to the founding lineages of Britain “

One cannot define reliably what the “founding lineages of Britain” are. This is a concept that is hard to define and even if it was defined properly, we probably could not assign one or several haplogroups to this population. This is another example of interpretative phylogeography, for which broad conclusions on population ancestry are drawn from allele frequencies in a limited number of modern populations.

“..we found people that have got Berber and Tuareg ancestry from the Saharan nomads”

Again, this is not something that can be concluded from the DNA data generated by Britain’s DNA. Perhaps AM means that he found people carrying a haplotype that is today present in the Berber and Tuareg, and no doubt also in other populations? Importantly, genetics of Berber and Tuaregs have been studied in more details than other populations, which can bias this inference. Because of the extensive migrations of human populations, we can only make guesses about where that ancestor was.

To summarize, human ancestry is a complex process with many unknowns. mtDNA and Y markers provide some interesting information about this ancestry, but nothing like what can be obtained from genome-wide data. But more importantly, one cannot link allele frequencies in modern population with definitive statements about where these alleles “originated”. This extrapolation cannot be supported by scientific studies, and therefore many of the claims made in the AM interview are not substantiated.

Lastly, going back to a comment I made earlier, I was surprised by AM’s statement: “we subsidize it massively”, when referring to Britain’s DNA product. As far as I know, such mtDNA and chromosome Y genotype tests can be bought from other businesses at a comparable, if not lower, prices. It therefore would have been helpful to explain what these subsidies are, and their exact extent. Is a grant agency providing some funding to support the company for example? This question matters because it is important to clarify the aims, even if commercial and scientific can sometimes be intertwined. There is nothing wrong about running a DTC genetic ancestry business of course, but one needs to state it clearly.


  • Digg
  • StumbleUpon
  • del.icio.us
  • Facebook
  • Twitter
  • Google Bookmarks
  • FriendFeed
  • Reddit

26 Responses to “Exaggerations and errors in the promotion of genetic ancestry testing”


  • I contacted Scotland’s DNA about some unsupported claims earlier this year after a series of credulous reports in the press. They stopped responding when I asked for citations.

    I also got grief on twitter from one of the journalists who wrote the story asserting that, because Moffat was a journalist of considerable reputation, how dare I impugn his business.

  • Thanks for sharing your concerns about the exaggerated claims made by this company. They seem to be very good at generating press coverage but not so good at providing accurate scientific information. We have a page in the ISOGG Wiki which provides details of the markers used in these tests.

    http://www.isogg.org/wiki/Britains_DNA

    As you will see the Britains DNA/Irelands DNA/Scotlands DNA test looks at just 200 Y-chromosome SNPs and 200 mitochondrial DNA SNPs. In contrast the new Geno 2.0 test from the Genographic Project tests 12,316 Y-SNPs and 3,352 mt SNPs:

    http://www.isogg.org/wiki/Genographic_Project

    The 23andMe test looks at around one million SNPs including around 2000 Y-SNPs and 2000 mt SNPs. Both these tests provide much more detailed information than the Britains DNA test and are much cheaper, though the 23andMe test is seemingly heavily subsidised by investor funding.

    Chip testing is not necessarily the best option for Y-DNA and mtDNA. For mitochondrial DNA it is now possible to get one’s entire mitochondrial genome sequenced (all 16,569 bases) at Family Tree DNA. This costs just $199 in the current sale, and will provide a detailed haplogroup assignment corresponding to the clades published in Build 14 of Phylotree. FTDNA customers can also participate in the mtDNA haplogroup projects which are run by volunteer project administrators:

    http://www.isogg.org/wiki/MtDNA_haplogroup_projects

    For males Y-SNP testing can be complemented with Y-STR testing at Family Tree DNA where one can join surname projects and haplogroup projects. The subclade can often be predicted from the Y-STR signature so that people can order a single SNP a la carte rather than having to pay out for a more expensive test. A list of haplogroup projects can be found here:

    http://www.isogg.org/wiki/Y-DNA_haplogroup_projects

    The “citizen scientists” who run these haplogroup projects usually know more about their haplogroups than the academics. They are currently mining the early Geno 2.0 Y-SNP results to determine the placement of the new SNPs on the Geno 2.0 chip on the Y-SNP tree.

    Many new discoveries in the Y-chromosome have come about through Family Tree DNA’s Walk through the Y programme:

    http://www.isogg.org/wiki/WTY

    We have charts in the ISOGG Wiki comparing the offerings from some of the major testing companies though the Genographic section needs to be updated:

    http://www.isogg.org/wiki/Y-DNA_testing_comparison_chart

    http://www.isogg.org/wiki/MtDNA_testing_comparison_chart

  • The “marker” that Moffat calls L1b is I believe the mitochondrial DNA haplogroup L1b which does occur in Africa but is also found at low frequencies in Europe. See: http://genome.cshlp.org/content/22/5/821.full There was a BBC story reporting on a study of African genomes which mentioned the Queen of Sheba in passing: http://www.bbc.co.uk/news/science-environment-18526428
    However, that is no reason to describe any haplogroup as being specifically “Sheban”. mtDNA haplogroup HV is found at low frequencies throughout the UK: http://www.eupedia.com/europe/european_mtdna_haplogroups_frequency.shtml

  • Naughtie seems to have taken the view that because it was science he couldn’t possibly be expected to probe

    btw, i listen to many bbc radio 4 podcasts as well as public radio bbc feeds here in the states, and this is a serious systematic problem i’ve noticed. many americans praise the bbc for their often skeptical/trenchant interviews as opposed to the soft-serve puffy stuff you get here in the states, but when it comes to science credulity seems to be the rule.

  • It would be good if this Group came down and appeared at WDYTYA. Also sometime in the next year or two there should be a 1-Day Symposium with all the companies that claim to participate in this area. It might be difficult to get everyone to travel, but they could appear on video-link these days in a suitable venue.

  • this interview sounds atrocious!

    By the way, I’m a bit bemused by your claim that Y chromosome and mitrochondrial DNA will only tell you about 2 of your 4,000 ancestors back 12 generations – it will also tell you about intermediate generations (or is that not included in your denominator of 4000?)

    Pretty poor that they’re resorting to legal threats, it’s like Ben Goldacre’s Bad Science sagas all over again – just from a science company instead of nutriceutical vendors (and involving the rector of a university no less, so much for the value of education…) I thought Britain had done something to sort out it’s ridiculous libel laws, at least as far as scientific claims were concerned?

  • Daniel MacArthur

    By the way, I’m a bit bemused by your claim that Y chromosome and mitrochondrial DNA will only tell you about 2 of your 4,000 ancestors back 12 generations – it will also tell you about intermediate generations (or is that not included in your denominator of 4000?)

    Vincent meant that of the 4,000 ancestors in that generation, mtDNA and chrY markers tell you about only two of them. Of course you’re right that these markers do also carry information about intermediate ancestors too.

  • Vincent Plagnol

    Thanks for replying to this Daniel.
    And Booker: yes I was simply computing 2^12 = 4,096, so looking only at the last generation.

  • @Booker:

    Pretty poor that they’re resorting to legal threats, it’s like Ben Goldacre’s Bad Science sagas all over again – just from a science company instead of nutriceutical vendors (and involving the rector of a university no less, so much for the value of education…) I thought Britain had done something to sort out it’s ridiculous libel laws, at least as far as scientific claims were concerned?

    Nah – this is still up in the air. See:

    http://www.senseaboutscience.org/pages/the-case-for-a-new-public-interest-defence.html

  • Far from the first uncritical interview of scientists I’ve heard on the BBC. It might be time to collate evidence and challenge them.
    On the 2e12 calculation, I know you were just using this as an illustration but we don’t really have that many ancestors or there’d be a lot more people in the past than in the present! How do pop genetics folk account for shared ancestry in their real calculations? I wonder what that number would really be on average?

  • It appears from this writeup about their new 2013 book, that AM likes to S T R E T C H and M A S S A G E the facts. See my comment below. As Mark Thomas has personally said to me via email, there’s a big difference between “Genetic Story Telling” and “Genetic Testing”. And I think Mark Tomas and James Wilson use to collaborative and write papers together … what’s going on with “that” relationship?

    New 2013 Book from AM & JW – Britain’s DNA: A People’s History

    Here’s what “they” write here http://www.alistairmoffat.co.uk/

    “Hidden inside all of us – every human being on Earth – is the story of our ancestry. Printed on our DNA are the origins of our lineages, the time in history and prehistory when they arose, and the epic journeys people have made across the globe.”

    “Based on exciting new research involving the most wide-ranging sampling of DNA ever made in Britain, Alistair Moffat, author of the bestselling The Scots: A Genetic Journey, shows how all of us who live on these islands are immigrants.”
    COMMENT: Does Maffat believe his minuscule DNA Testing Company has a WIDER RANGER of testing / sampling than the POBI Team who has spent US$=5 million plus?

    “The last ice age erased any trace of more ancient inhabitants, and the ancestors of everyone who now lives in Britain came here after the glaciers retreated and the land greened once more. In an epic narrative, sometimes moving, sometimes astonishing, always revealing, Moffat writes an entirely new history of Britain. Instead of the usual parade of the usual suspects – kings, queens, saints, warriors and the notorious – this is a people’s history, a narrative made from stories only DNA can tell which offers insights into who we are and where we come from.”

  • @Mike. As you note the calculation of the number of genealogical ancestors as 2^k k generations back quickly reaches then exceeds the population size. What is happening there is that individuals within the population are your ancestor multiple times over. I.e. an individual a number of generations back can be traced back to being your ancestor through two distinct paths back through your family tree. In a population of size N individuals it takes roughly log2(N) to reach the point where multiple paths through the pedigree start to point to the same ancestor (so if the population was ~10 million roughly 23 generations, although we should probably use population size of a specific region and so reach this point sooner). When an ancestor can be reached by multiple routes through a family tree of a present day individual we call this an inbreeding loop, as the individuals parents were related. Due to the fact that our number of possible ancestors grows quickly to the population size we are all mildly inbred.

  • Vincent Plagnol

    Mike, Razib: thank you both for your comments, and Razib for your follow-up blog post.

    One bit of information about the interview relevant to the points you made (and which I only learnt today) is that the BBC host J. Naughtie is in fact a good friend (I am quoting Naughtie here, see link below) of Alistair Moffat. There is obviously nothing wrong about inviting a friend, but based on the recording it was not stated in the interview. In any case this link, where J. Naughtie expresses his support for A. Moffat as rector of St Andrews, probably explains at least in part the friendly tone of the interview:
    http://www.youtube.com/watch?v=_1mID06KTCM

  • That really is outrageous. Its a blatant advert for the company of a friend presented uncritically as leading edge science. Even if Naughtie is gullible enough to believe this stuff, the editors should have stepped in with some real questions about evidence and peer acceptance and questioned the commercial aspect of the work. The BBC is not in a good place at the moment. They have been crucified by a parliamentary committee and their editorial standards have been called chaotic and totally inadequate. Yet no-one gets sacked – the worst of them get huge pay-offs at our expense. One wonders what interests lie behind many of the other questionable reports that crop up on BBC news. Shocking.

  • David Balding

    Vincent, Mike: I think this is one of the most shocking aspects of the story. Jim Naughtie has interviewed his friend Moffat more than once on BBC radio, without to my knowledge the friendship connection being mentioned. Moffat seems to have little relevant expertise, and has taken the opportunity to promote his business interests without making it clear that there was a business interest involved. From what I’ve heard Naughtie has not asked his friend any kind of probing question, and no qualified scientist was invited to comment on Moffat’s dubious claims. And when I complained to the BBC about this, the complaint was dismissed with no real response to the issues raised (I wasn’t aware of the friendship at the time of the complaint).

  • The Sunday Times has picked up this story with a big article today on page 3 of the main newspaper. The article begins:

    “WHEN James Naughtie presented a gentle item on the Radio 4 Today programme about tracing ancestors through DNA testing, it hardly seemed likely to cause a stir.

    Instead it has plunged the veteran broadcaster into a row involving warring academics and the Queen of Sheba.

    In July, Naughtie interviewed his old friend Alistair Moffat, whose business interests include BritainsDNA, which traces the “deep ancestry” of individuals through a saliva test. The relationship between the two men was not revealed to listeners.”

    The full story is behind a paywall here:

    http://www.thesundaytimes.co.uk/sto/news/uk_news/Tech/article1182305.ece

  • I have been tested by FTdna and also by Genebase in Canada.

    As a result it has transpired that I have a close(ish) connection with a number of McCrackens. The number of markers tested was 67. 65 matching.the unmatched markers were CDY a & b. I read that some people ignore these markers as being too volatile..My own results for these markers 39 – 41 is pretty rare. I made a comparison with other members of an FTdna group and it appeared to have an occurrence of under 1%. I asked FTdna if they would tell me the official occurrence figures for specific scores within markers and they refused, which I found extraordinary.

    my ‘score’ for these markers with Genebase was 39-40… Genebase said they knew of no reason why the marks should differ. FTdna said they retested and stood by my original result….It leaves me perplexed…

    Why is 23andMe more useful, and could it be of any value to be included in a ‘different’ and wider database if that would indeed be the case.
    Regards Mel Wibberley

  • Mel

    You can see a list of frequencies of the values for the different markers on this website:

    http://freepages.genealogy.rootsweb.ancestry.com/~geneticgenealogy/yfreq.htm

    Sometimes different companies report different values for different markers because there are different reporting conventions, though I don’t believe CDYa/b is one of the markers affected. At some point all companies should convert to the NIST standards. There is no reason to doubt the FTDNA result. CDY is the most volatile of all the markers.

    The Y-DNA test you’ve taken with FTDNA is a genealogical DNA test and gives you matches with people you are related to in a genealogical timeframe on the direct paternal line which normally coincides with the transmission of surnames. 23andMe has a Relative Finder feature which gives you matches with your genetic cousins on all your family lines but you won’t know which line the match is on. The 23andMe test does include around 2000 Y-chromosome SNPs. These are used for deep ancestry connections and cannot be used for surname matching. I would suggest you have a look at some of the reviews that are linked on the 23andMe page of the ISOGG Wiki to understand how the test works:

    http://www.isogg.org/wiki/23andMe

  • Paul O'Reilly

    That interview was laughably ludicrous. Alistair Moffat sounded more like a 2nd-hand car salesman than the main public figure of an academic institution – aren’t St. Andrews embarrassed by this? (the BBC should be too, but have presumably developed a pretty high shame threshold lately). Good to see this so well exposed by this article.. just a couple of other points:

    I think any talk of Adam & Eve here is misleading – the concept of mitochondrial Eve & a chr-Y Adam, is about as meaningful as a chromosome7,basepair-1.5Mb ‘Alex’ … ie. we’ve got millions of most recent common ancestors corresponding to millions of loci across the genome..there’s not much more reason to focus on these two ancestors than any of the others.

    Pretty much every word that Moffat said during that interview was plucked from the land of make-belief of course, but one point that hasn’t really been highlighted was [what appeared to be] the cheeky attempt to claim that just because there had only been a couple of mutations at a single marker since ‘mitochondrial Eve’, then that somehow implied that “Eve’s present-day grandson”[!!] had essentially the same DNA across the entire genome as “Eve” ['very pure' as naughty Naughtie said] – by that logic we’re all the direct grandchildren of the pre-historic ancestors common to the Chimpanzee & ourselves, since we’ve got *exactly* the same allele (0 mutations since..) at most of our genome as the Chimp (most of us, at most genomic positions).

  • Mr. Moffat seems content to engulf himself in the “Mist and Low Cloud” obscuring of history that Neil Oliver refers to in the “History of Scotland” series.

    It is not becoming to the institution of St. Andrews, the Kingdom of Fife, or the People of Scotland, to engage in profiteering at the expense of scientific and commercial honesty.

  • There’s another radio interview with Alistair Moffat in a similar vein which might also be of interest:

    http://www.bbc.co.uk/programmes/p00r6w16

    An article was also published in the Scotsman:

    http://www.scotsman.com/the-scotsman/scotland/scotland-s-dna-descended-from-lost-tribes-and-related-to-napoleon-1-2238030

    As far as I’m aware no research has as yet been published in any peer-reviewed scientific journal.

  • 4000~ 2^12 DIFFERENT ancestors?
    so no cousin marriages, not even 11th cousins
    c’mon

    what was the population 300 years ago if we’ve each got 4000 ancestors?

    although obviously the claims in the radio4 broadcast are are unscientific & way below the standard we should expect from BBS from any programme

  • @diana

    To be fair, Vincent did say “about 4,000″, not “exactly 4,092″. I don’t really know much about the subject, but Kenneth Wachter estimated that amongst British people 5% of direct ancestors at 15 generations are “duplicates” – so we are talking about having something like 3900 ancestors at 12 generations back. “About 4,000″ sounds about right then…

  • Well done Genomes Unzipped, I’d call that a victory for common sense.
    http://www.bbc.co.uk/news/science-environment-21687013

  • My partner bought this test for me as a Christmas gift (male & female line). I was informed that my 1st sample failed and was asked to provide another, the 2nd also failed and so did the 3rd. I was then told 3 months later they can’t provide another test due to cost reasons. I asked for some explanation from the lab as I was meticulous with following their instructions, none was provided and I was refunded the money – minus obviously all the postage costs for 3 special delivery packages.

    Overall, I’m very disappointed with their service and certainly would not recommend.

    Graeme

Comments are currently closed.

Page optimized by WP Minify WordPress Plugin