As regular readers will know, I am comparatively new to the convoluted ‘excitements’ of Ancestry DNA matches. I’ve been having a little play. If you have nothing better to do, you might see how your matches compare. You probably won’t all read to the end so I will put today’s new novel hint at the beginning – #5 It contains a modern strand. The novel also has the beginnings of its own webpage now. This post is probably for those who like fiddling with numbers and charts. If that’s not you, please go and use your time more wisely and stop reading now.

So, back to the DNA. I have, according to Ancestry, two 3-4^th cousin matches, 316 4^th-6^th cousin matches and as of this minute, 35,218 5^th-8^th cousin matches. I have looked at all those with common ancestors (allegedly 86 people although a couple have some incorrect trees). I have also checked all those that match at 16cM or more and all the shared matches of these people, plus a random selection of others. That’s over 2000 matches in all. So far, I have identified precisely how I am related to 78 people. The DNA that I share with some of these individuals is a little as 6cm. I am also more closely related to some than ancestry believes. Thanks to shared matching, I have an additional 237 matches, where I can tell which branch of the family is likely to be responsible for the link, although not the precise details of how we are connected.

I am very far from being a DNA expert, so please do tell me if all this is total nonsense. Anyway, I thought it might be fun to consider how my matches are distributed amongst different branches of my family and to speculate on why not all my great great grandparents are equally represented.

My paternal grandfather’s paternal grandfather

The Braund family from Devon and Cornwall can be traced back to the 1680s. 28 of my identified matches relate to this branch of the family and I know precisely how 20 of them relate to me. The closest relations are two 3^rd cousins once removed, with whom I share 35 and 20cM. Most are 6^th cousins but I do have a 7^th cousin (who matches at 11cM) and a 7^th cousin once removed (with 15 shared cM). I have not ruled out being related twice over with some of these individuals. In fact, in one case, I know I am.

My paternal grandfather’s paternal grandmother

The Nicholls family, also from Cornwall, can be traced back the 1630s. They yield just six matches, only one of which I can precisely identify. This is a 4^th cousin once removed (a 20cM match).

My paternal grandfather’s maternal grandfather

The Bishops, another west country family, are responsible for a whopping 104 matches, most of whom have no trees or private trees. The 12 for whom I have precise details of how we are related, vary from a 3^rd cousin once removed, who only shares 12cM of DNA, to several 6^th cousins, one of whom shares 14cM.

My paternal grandfather’s maternal grandmother

The Buckinghams and their forebears come from Cornwall. I have 39 matches that I have identified as relating to this branch, three of whom have a known place on the family tree. These range from a 5^th cousin (21cm) to a 6^th cousin (11cm). The closest relative shares 43cM of DNA (precise relationship unknown).

This quarter of my family is responsible for nearly 60% of my identified matches. There are obviously a number of factors at work here. Firstly, as I have not inherited an equal amount of DNA from each grandparent, I am less likely to have matches with those whose DNA is more diluted in my make-up. Perhaps I have a higher proportion of west-country DNA (although Living DNA’s ethnicity estimate suggests not). In addition and probably more significantly, the south-west branches are likely to be linked to a higher proportion of emigrants, therefore there will be more residents of the USA amongst the wider family. Given the much higher number of US citizens who have tested, this is bound to have an impact. Also, by using shared matches to identify likely group members, there is a snowball effect.

My paternal grandmother’s paternal grandparents

What about the other three quarters of my ancestry? My paternal grandmother’s quarter is the poorest yield for DNA matches, with none for her father, the Hoggs from Northumberland and just two 4^th cousins (19cM shared) for her paternal grandmother. I believe that this is largely because these lines are more difficult to trace, not just for me but for other researchers too. The Hogg line hits a brick wall in 1804. Therefore matches on these lines are probably hidden within the many hundreds of matches for whom I cannot identify a common ancestor. Of course, I cannot ignore the possibility that Mr Hogg may not actually be my genetic great great grandparent.

My paternal grandmother’s maternal grandparents

The Howe and Stratford lines from Buckinghamshire (with 7 matches, 4 of which are identified), give me one of my closest matches, a third cousin, yet we share just 27cM. It looks as if this grandparent is under-represented in my DNA, another possible reason for fewer matches. So, fewer than 3% of my matches come from my paternal grandmother.

My maternal grandfather’s ancestors

On my mother’s father’s side we hit the problem of endogamy; with the Smiths and the Seears intermarrying in three successive generations. This means that I am related to most of my matches in more than one way, making meaningful analysis difficult. I do have 46 matches in this line, five of whom can be located on my family tree. I have one matche to my Norfolk great great grandparent, Anne Bulley, a 4^th cousin sharing 18cM of DNA). This again is a difficult-to-trace branch. Despite this, 15% of my matches relate to my maternal grandfather.

My maternal grandmother’s ancestors

Finally, to my maternal grandmother’s line. The Sussex Woolgars are well documented and can be traced back to the C15th, so perhaps it is not surprising that there are 23 relatives on this branch, ranging from a 4^th cousin (with only 6cM shared) to a 6^th cousin sharing 17cM. Six of these can be precisely identified.

On the virtually impossible to trace Cardell line I have two matches. These support my speculation as to the ancestry of my brick wall ancestor but more is needed.

The Essex Dawsons and Bowyers can be traced back to the 1650s, again giving potential for plenty of matches and there are 47 of them, of which I can pinpoint the exact relationship for 16, all of whom are 4^th– 6^th cousins. Thus, this grandparent gives me 23% of my matches, not far from her fair share.

In summary then, this is the percentage of matches that each grandparents is responsible for:-

Paternal grandfather 60%, paternal grandmother 3%, maternal grandfather 15%, maternal grandmother 23% – sorry, I know that adds up to 101% – blame rounding up/down.

Ancestry have just announced their Mother’s Day sale. Shall  take the plunge and get a daughter to test, so I can do all this for her paternal ancestry as well?

* My only connection with any DNA testing company, is as a customer. I have received no concessions, free gifts or financial inducements from any of them.

And just because I can, one of my favourite family photos, colourised by MyHeritage and then enhanced by me.

Well, that was exciting. My updated ethnicity estimate from Living DNA is in. My original results, which I received in June 2017, were a pretty fair reflection of my  what I expected, based on my documentary tree. Compared to my estimates from other companies, it was the closest match to what I have discovered during over forty years of research. Here is what I wrote at the time, some of which I have repeated here.

Today’s revised estimate is based on Living DNA’s improved, more refined data. There are no massive changes but it is interesting to examine the subtle differences and how the new results compare to my predictions, based on the paper trail.

I am now in my fifth decade of family history research and have an extensive pedigree based on documentary research. I am fortunate to know the geographical origins of 31 of my 32 3 x great grand-parents and 75% of the generation before that. This takes me back to the mid-eighteenth century, well before the point at which my ancestors converged on Greater London. Of course, these estimates are just that, ESTIMATES but as the science improves, we might expect that they will become more accurate.

In preparation for the original results, I used my research to work out what I might expect. I am aware that the DNA that I have inherited does not come equally from all my 3 x great-grandparents and that some of them may have left no trace in my profile but I had no way of taking account of this. I am also assuming that my documentary pedigree is correct. Matches at other testing companies have, so far, not given me any reason to think that my genetic tree is different from my paper one. Based on my knowledge, my expectation was that my genetic make-up would reveal:-

Cornwall 20%

Devon 5%

Northumberland 12.5%

South Eastern England 37.5%

South Central England 19%

East Anglia 6%

So, as I did with the first results, let us examine a region at a time. Firstly the south-west, Devon and Cornwall. By my reckoning, the south-west makes up 25% of my ancestry, with 20% being Cornish and 5% for Devon. Living DNA’s first percentages were 7.4% Cornish and 11.7% from Devon. Now, the overall percentage is the same – 19.1% but the distinction between Devon and Cornwall has become less accurate, rather than more, with 3.5% Cornish and 15.6% now being allocated to Devon. As my south-west ancestors lived very close to the Devon-Cornwall border, on one side or another, I am not disappointed with this.

Next, Northumberland, which I would expect to be 12.5% of my ancestry. With the original results, by adding the estimates for Northumberland (5.8%) and Cumberland (7.2%), I had the anticipated 13%. This segment of my ancestry has now been refined to be slightly more accurate, with 7.9% for Northumberland, 1.1% for the North-west, 1.9% N.Ireland/S.W. Scotland and 1.5% Aberdeenshire, a total of 12.4%. I suspect that the lost 0.6% has defected to Europe.

In the south-east, again the more refined breakdowns have become less similar to the documentary tree. I believe that the bulk of my ancestors, 37.5%, come from the south-east, Essex and Sussex. Last time, Living DNA agreed, with 35.3%. Now, the new results show only 7.9% of my make-up being from the south-east. I would anticipate a further 19% to come from the south-central region. This was under-represented first time round, at only 3.9%. The swing to the south-central region has been huge; I now have 33.7% from there. Living DNA has also made minor adjustments to the percentages from the south, which has moved from 5.8% to 4.2% and to the central region, which has gone from 2.7% to 1.4%. So once again, I find that the percentage for the general area is what I would expect but the distribution within that area has become slightly less accurate.

So what does that leave? There is still no sign of my expected 6% East Anglian ancestry. As I commented last time, I suspect that that has become Scandinavian ancestry, which has decreased from 5.6% to 4%. Lincolnshire has increased from 1.2% to 2.1%. I have lost the anomalous 2.1% from Chechnya and gained 7.5% from north-west Germany, perhaps reflecting early migrations to the east coast.

What fascinated me last time and still does, is the alleged Yorkshire ancestry, which has decreased slightly, going from 11.1% for North Yorkshire, to 6.4% for North Yorkshire and 1.2% for South Yorkshire.  At present, I have one great, great grandmother whose origins are unknown but my best guess would be the Shropshire/Worcestershire borders. Should I be looking at Yorkshire instead, or might this be the DNA manifestation of Shropshire/Worcestershire?

I also have a 4 x great grandmother who appears on my tree in that position 3 times, due to the marriage of cousins in two successive generations, thus she represents 4.7% of my ancestry. I do not know her surname and I have no idea of her origins. At present, I have assumed she was from the south-east, which is where her children were born in the 1780s; I wonder if I may be wrong.

So, overall, I still find that Living DNA’s regional breakdown is broadly similar to my documentary tree. The minor tweaks that have come with new results have been a case of win some lose some. Overall though, I am pleased with Living DNA. I am still waiting for the updates on two other kits, so I will be reporting back further then. I am especially interested to see if the one which diverges dramatically from the paper trail has changed.

Map created using Genmap

* My only connection with Living DNA, or any other DNA testing company, is as a customer. I have received no concessions, free gifts or financial inducements from any of them.