Saturday, 5 December 2015

My thoughts on DNA Cymru Part 3 and the significance (or lack thereof) of large genetic clusters

The third part in the controversial S4C series DNA Cymru was broadcast on the Welsh-language TV channel S4C last Sunday. The programme (with English subtitles) is available on the BBC iPlayer for the next 28 days, and can also be viewed on the S4C catch-up website.  For background read my reviews of Part 1 and Part 2.

The DNA Cymru series has been criticised because of the perceived use of a public service broadcaster to promote a commercial company in the guise of a research project. In order to participate in the "research" it is necessary to purchase an expensive DNA test, and the TV programmes themselves have been used to promote the sale of the kits. The results of the "research" have not been published in a scientific journal, and have therefore not gone through the usual process of peer review which is the minimal standard for credibility.

In the third programme in the series we were presented with further findings from DNA Cymru's "research". Dr Jim Wilson, the Chief Scientific Officer of Cymru DNA Wales, featured prominently on the programme. CymruDNAWales is one of a family of companies operated by the Moffat Partnership.* Jim Wilson is a major shareholder of this company, though his financial interests were not mentioned in the programme. CymruDNAWales have partnered with S4C to produce the DNA Cymru series. It also turns out that Alistair Moffat, the co-founder of the Moffat Partnership, is an old friend of Ian Jones, the CEO of S4C.

Chieftains, noblemen, princes and kings
The main finding of this episode was that "a high percentage of today's population are the descendants of brave warriors alluded to in our national anthem  the nation's ancient chiefs and princes". This "research" is based on the test results of 1000 people who have paid for a Y-chromosome DNA test from the CymruDNAWales website over the past year. A Y-chromosome test is taken by males, and focuses on the patriline - the line of inheritance which usually corresponds with the transmission of surnames. No details were provided as to the qualifying criteria for inclusion in the study other than an ability to pay for a DNA test, though I imagine that participants would probably be required to have a paternal grandfather who was born in Wales.

An analysis of the Y-DNA results revealed that there were twenty genetic clusters in Wales. These clusters were defined as "unique sub-branches in the Y-chromosome tree". Ten of these clusters were analysed in more detail to get an estimate of their age, and in some cases this involved sequencing the Y-chromosome.

The claim was made that these ten genetic clusters trace back to ten men who were born about 1000 to 2000 years ago, and that these men are the direct male-line ancestors of 18% of all Welshmen today or around one in five of the male population. This figure equates to over 200,000 men in Wales and nearly 500,000 men all around the world. It was further claimed that these ten men were probably all brave warriors, prominent chieftains, noblemen, princes and kings who had "shaped society in a way that no one had suspected before". Dr Jim Wilson commented: "They certainly outbred their peers. Probably they were nobleman, chieftains and kings. If this prominence is inherited the ability to have more sons is inherited down the male line."

It was stated that similar genetic clusters had been found in Scotland for some of the Scottish clans and in Ireland where many males are the descendants of early chieftains, though the clustering seen in Wales was more extreme than that seen in Ireland and Scotland.

The details were not given on the programme but there was a study carried out in 2006 out by a team at Trinity College Dublin which found a prevalent Y-chromosome DNA signature in Ireland (Moore et al 2006). It was suggested that this signature showed "a significant association with surnames purported to have descended from the most important and enduring dynasty of early medieval Ireland, the Uí Néill ". This line traces back to a "possibly mythological 5th-century warlord" known as Niall of the Nine Hostages. However, the testing in this study was done at very low resolution (17 Y-STRs), and there is considerable uncertainty over the dating of the TMRCA (time to the recent common ancestor). Better TMRCA estimates can now be obtained from the results from the newer next-generation sequencing tests. The findings have also been criticised on historical grounds (Swift 2013).

No equivalent study has been carried out in Scotland, though there are numerous surname DNA projects run by genealogists investigating Scottish surnames. Dr Jim Wilson mentioned on the programme that a large number of people with the surnames MacGregor and Stuart formed a big genetic family - a genetic cluster of closely related Y-chromosomes. There are large DNA projects for the McGregors and the Stewarts. Both projects are hosted at Family Tree DNA, but some of the project members will have had supplementary testing through one of the Moffat Partnership companies.

For an independent opinion on the question of the plausibility of the suggestion that the "genes of people today" can be connected with "the bloodline of historical princes" the programme paid a visit to Dr Chris Tyler-Smith at the Sanger Institute in Cambridge. Tyler-Smith's lab published a paper in 2003 on "The genetic legacy of the Mongols" (Zerjal 2003), and he described the results of this research. A large Y-chromosome cluster   a set of closely related Y-chromosome types  – was found purely by chance in the "genetic landscape across Europe and Asia". The cluster was thought to date back to Mongolia about 1000 years ago. This large cluster could be a result of natural selection where "the lineage has a particular biological advantage" but that is not thought to be a plausible explanation in the case of the Y-chromosome. The alternative is that there is "some social advantage. In simple terms this means that one man has fathered many different children, and this social advantage must have lasted over many generations." It was suggested that the most likely explanation for this large genetic cluster is that it is linked to "the exploits of Genghis Khan" who is thought to have around 20 million living descendants.

Tyler-Smith also appears to have been given the opportunity to look at the data from the DNA Cymru project. He commented that the number of clusters was remarkable, and he also noted how much of the population was included in these clusters which was higher in the Welsh data than in other datasets he'd seen.

The programme's researchers seemed to be unaware of a subsequent study by Balaresque et al (2015) which analysed the Y-DNA signatures of 5321 males in Asia and identified 11 descent clusters "that represent likely past instances of high male reproductive success". These clusters were associated with "expansions that began between 2100 BCE and 1100 CE, found both among sedentary agriculturalists and pastoral nomads". One of the clusters corresponded with the "Khan" descent cluster identified in the Zerjal study. The authors commented that this was "the most striking signal of an Asian expansion lineage, representing 2.7% of the entire data set, and the highest number of identical Y-chromosomes (N=71)". They suggested that "High reproductive success is often associated with high social status, ‘prestigious’ men having higher intramarital fertility, lower offspring mortality and access to a greater than average number of wives". They also suggested that with next generation sequencing and new ancient DNA approaches it might be possible to identify "the prestigious and powerful pastoralist founders" of these high-frequency descent clusters.

But how plausible is the suggestion that large descent clusters must always be associated with prestigious and powerful people? Is there another explanation? Could such clusters occur by chance? Guillot and Cox (2015) have published an interesting preprint in response to the Balaresque paper which investigates this question using computer simulations. The paper is currently going through an open peer review process. The authors concluded:
The most parsimonious explanation is therefore that the high frequency haplotypes observed in Central Asia are simply expected chance events, and an explanation invoking cultural transmission of reproductive success is not necessary to account for them. As no other evidence is presented to support proposed links to famous historical men, these haplotypes instead most likely reflect the chance proliferation of random male lines, probably from historically unrecorded, culturally undistinguished, but biologically lucky Central Asian men.
Similar conclusions were reached in a study of 131,060 Icelanders which combined genetic information with detailed genealogical records (Helgason et al 2003). They found skewed distributions of both mtDNA and Y-chromosome haplotypes:
These results demonstrate even over a short timeframe of a few hundred years, during which the size of the Icelandic population increased almost fivefold, only a minority of potential ancestors actually contributed mtDNA or Y chromosomes to the contemporary population. The vast majority of contemporary females (58,832 or 91.7%) are descended from only 22% (7,041) of the potential matrilineal ancestors born between 1848 and 1892, and most contemporary males (57,686 or 86.2%) are descended from only 26% (8,275) of the potential patrilineal ancestors. The results are even more striking for matrilines and patrilines traced back to the 1698–1742 ancestor cohort.
It may well be that with ancient DNA samples and further genealogical research it will be possible to attribute some high-frequency haplotypes to specific prestigious individuals, but it seems plausible that many of these clusters have arisen by chance and it will never be possible to identify the founders.

The Tudors
The programme also included a discussion of the results from a rather puzzling study of the Tudor surname. Men with the surname Tudor were invited to take a Y-DNA test as part of the "project" in an attempt to identify the genetic signature of the royal Tudor lineage. However, such an investigation was doomed from the start because there are no known male-line descendants of Henry VII and Henry VIII. In addition, Wales has traditionally used a patronymic naming system so there is no historical link between the surname and the Y-chromosome. The adoption of hereditary surnames was a gradual process, and the patronymic system persisted in some parts of Wales right through until the nineteenth century. The undisclosed number of men who had their DNA tested as part of the Tudor research were divided into three genetic families, and the results were not surprisingly described as "inconclusive".

However, although there are no direct male-line descendants of Henry Tudor it is possible to trace his tree further back in time to Ednefyd Fychan, a prominent 12th-century Welsh leader. The line can then be traced forwards through a different line of descent to the present day. Two males descended from Ednefyd Fychan were identified and had their DNA tested but the DNA results did not match. However, if further descendants can be traced then it may well be possible to determine the Y-DNA signature of the Tudors.

Some genealogy
The rest of the programme was taken up with an investigation of the genealogical lineages of some of the men who had taken part in the DNA Cymru "project" with a particular focus on noble and ancient lineages. While the emphasis on genealogy was most welcome the approach seemed somewhat ad hoc, with the DNA testing done first and the genealogy explored afterwards but in a very unstructured way. In each case only one person from the lineage was tested, whereas the usual practice in a genealogical research project is to test at least two cousins who share descent from the ancestor in question in order to identify his ancestral haplotype (DNA signature) and to rule out the possibility of what is known as a non-paternity event. I'm not familiar with the ancient Welsh genealogies but it also seemed that many of the ancient connections were highly speculative. We would not normally expect genealogical research to go through the peer review process, but we would hope to see the research published on a website or in a respectable genealogical publication such as The Genealogist, the journal published by the Society of Genealogists. None of the research featured on the programme appears to have been published anywhere so I've summarised the results briefly below.

A man by the name of Jim Williams was interviewed. His family had lived in Llywele as far back as the 15th century, and the family claim to be descended from the Welsh king Hywel Dda. His haplogroup was R1b-S300 (L371)

Jim Cockburn-Powell has a lineage which can be traced back to one Roland Powell who was born in 1599. His male-line ancestry can then possibly be linked with a man called David Powell of Flint who appears in the Welsh genealogies and this line ultimately traces back to Cunedda Wledig, a British chieftain of the Votadini in south-east Scotland. Jim Cockburn-Powell belongs to haplogroup R1b-PF5191 which we were told is rare in England and not seen outside Wales.

Alexander Talbot-Rice lives at Newton House, the home of the Barons of Dinefwr, and is supposedly descended in the direct male line from a George Rice who married Lady Cecil Dinefwr, Baroness Dinfwr. George Rice is descended from Henry Rice Fitz Urien whose lineage supposedly goes all the way back to Urien Rheged, a sixth-century king. Alexander Talbot-Rice's haplogroup is G-Z674 which is apparently "highly enriched" in Wales, and particularly in southern Wales.

At the end of the programme we were introduced to the Welsh rugby player Colin Charvis. His dad is from Jamaica and his Y-line can be traced back to his great-great-grandfather Frederick Charvis who was born soon after the abolition of slavery in Jamaica. He would be expected to have an African Y-DNA haplogroup but his results showed that he is R1b-A228, a genetic cluster which is found predominantly in Wales.

Genetic genealogy  the use of DNA testing in combination with genealogical and historical records  is a legitimate and credible process. It was therefore encouraging to see a greater emphasis on genealogy in this programme, though it would have been interesting to have had feedback from historians and expert family historians who could have commented on the credibility of the published Welsh genealogies. While stories about kings and nobles are good for attracting the headlines and publicising sales of DNA kits, unfortunately the claims made on the programme were pure speculation. In any case if you want to make a programme about the DNA of Wales you need to look at the entire population rather than focus on celebrities and the elite classes. This episode focused solely on the results of Y-chromosome testing which by definition excluded half the population of Wales. I hope this means that a future programme will focus exclusively on the DNA of the female lineages of Wales.


Balaresque P et al (2015). Y-chromosome descent clusters and male differential reproductive success: young lineage expansions dominate Asian pastoral nomadic populations. European Journal of Human Genetics 23, 1413-1422.

Guillot EG and Cox MP (2015).  High frequency haplotypes are expected events, not historical figures. Preprint. BioRxiv. Published online on 8 July 2015.

Helgason et al (2003). A populationwide coalescent analysis of Icelandic matrilineal and patrilineal genealogies: evidence for a faster evolutionary rate of mtDNA lineages than Y chromosomes. American Journal of Human Genetics 72(6): 1370-1388.

Moore LT et al (2006). A Y-chromosome signature of hegemony in Gaelic Ireland A Y-chromosome signature of hegemony in Gaelic Ireland. American Journal of Human Genetics 78(2): 334-338.

Swift C (2013). Interlaced scholarship: genealogy and genetics in twenty-first century Ireland. In: Duffy S (ed). Princes, Prelates and Poets in Medieval Ireland. Four Courts Press.

Zerjal (2003). The genetic legacy of the Mongols. American Journal of Human Genetics 72(3): 717-721.

*The other trading names of the Moffat Partnership are BritainsDNA, ScotlandsDNA, IrelandsDNA and YorkshiresDNA.

© 2015 Debbie Kennett