MacGregor Project Blog 2022

Welcome to my annual blog for the MacGregor DNA project. This past year has not seen a great deal of development in the relationship between genealogy and DNA as far as the project is concerned, although it has been good to observe that people have used the opportunity of being at home during the pandemic to do more genealogy, and to order Y700 tests which will, during next year, as the results begin to gather, potentially allow for more focused research on the various family  relationships which are being revealed by the SNP links. 

     To begin this discussion I thought it would be useful to show how SNPs can be used to suggest origins, not just genetic, but geographic. This year, therefore, I am presenting case study based on the McAdam, Macpherson, and Irish results. The second half of this blog will consist of an update on Gregg (Greig/Gragg) and Gregory SNPs, and conclude with comments on the Rob Roy descendant discussion, following on from last year’s blog.

                                McPherson/McAdam

During the year, I was contacted by a McPherson participant who could not work out why in his Y700 results his nearest matches were McAdams. The genealogical origins of the two surnames are quite different. Macpherson as a surname is mostly associated with a group of families originating in the south of Inverness-shire – round the areas of Kingussie, Newtonmore, Insch and Laggan (see this parish map for location http://www.scotlandsfamily.com/parish-map-inverness.htm). The McAdams on the other hand mostly originated in south west Scotland below the central belt, more or less the opposite end of the country. They also were found in Ireland and there is a fairly direct connection between Southwest Scotland and Ireland.

     If we look at the spider diagram for the McAdams [also Adam, McAdam and McCadam] (Figure 1) who are in the MacGregor Project we can see that there is a very large group, and then several apparently unrelated singles (such as 499181) or pairs (such as 510026 and 49834) results.

Figure 1: McAdam/Adam/McCadam

Next, in Figure 2, the kit numbers for those results which seem to suggest an Irish origin, which, in a good number of cases, is confirmed by known genealogy (plus IN100960 McPherson)

Figure 2:  McAdam/Adam/McCadam group connected with Ireland (bottom right)

Figure 3 shows the spider chart for the Irish group (all the charts are based on 67 markers):

Figure 3: the Irish group (kit numbers only)

In Figure 4 I have overlaid those McAdam results onto the Irish results and included the kit numbers with the addition of the first letters of the surnames (this being all the program will allow me to do). However, you will see that the McAdams in that tight group of results is embedded within the Irish results – for example kit 165907 or kit 183682. This suggests that these surname groups arose in the same location but in different individuals, creating strong genetic relationships which are not indicated by the surnames. You can see that the Irish group, which appears to have three separate branches coming from the main stem, contains various surnames including White, Black, and, what is especially notable, a tight and separate group of Grier/Greers.

Figure 4: McAdam results and McPherson IN100960 embedded in the Irish group

One thing I do need to mention at this stage is the limitation of STR testing and subsequent mapping of relationships. Since STR mutations are random, comparisons of results can be influenced by the occurrence of a mutation earlier in the sequence of 37, 67 or 111 markers – in other words if a mutation occurs in the first 25 numbers as read from left to right the program assumes that this is a significant mutation, because, in the early days of DNA testing for genealogy certain markers were chosen because they mutated more rarely than others.  With the program assuming that a mutation in this area is more significant it is possible for an individual’s results to be grouped with others which have that same mutation rather than the program ignoring that and focusing on later mutations so occasion misalignments may take place.  This limitation alone makes the use of terminal SNP testing more reliable for grouping individuals together (and why Y700 uses the SNP tree format designed by Alex Williamson and which Family Tree DNA call Block Tree). I discuss the meaning of ‘terminal SNP’ later (see the subsection heading below).

    Now to return to the question which I began with. In Figure 3 kit number IN100960 is McPherson and it is quite clearly in the same group as the McAdams but the line is longer indicating greater predicted passage of time since the shared ancestor. The most obvious interpretation of this result is that IN100960’s ancestor originally came from Ireland pre surnames. It is said that Ireland seems to have had the earliest use of surnames but by the 12^th century it was only really the highest class who would be using them, while most of the population would be known by patronymics (e.g. John son of Patrick son of John). The Ancestry blog website https://blogs.ancestry.com/cm/history-of-irish-surnames-is-yours-here/ says the following:

‘It was around the 1100s, as the population was increasing, that people in the upper social classes started taking hereditary surnames (those that remain fixed over the generations); others didn’t need surnames, or even get around to them, until the 1500s’

     Perhaps the earliest that that IN100960’s ancestor could have come over to Scotland would have been in the settlements that took place not long after 500AD, but it could equally be much later especially since the distance between Ireland and the Mull of Kintyre in Scotland is just 15 miles across the North Channel between the two land masses.  Over several centuries one can imagine the movement of people who were originally settled in Argyllshire migrating further north to Inverness-shire, so that one of IN100960’s ancestors found himself in the lands of the MacPhersons, and, instead of using patronymics started to use the surname of the local landowners.

     What I have just suggested follows the DNA and social history and is perfectly possible as an interpretation of IN100960’s DNA result, but it should be noted that it is only one interpretation. There are still comparatively few individuals who have DNA tested to find their terminal SNP but of those that are available, and checking not just the terminal SNP, but also those which came immediately before it shows that there are related individuals in Scotland and Ireland. It is possible to look at results which have been mapped onto a  very useful SNP mapping website:

https://www.genetichomeland.com/homeland.asp. You have to register to use this site but once registered (it’s free) you can map the results for most SNPs which gives you geographic locations for these SNPs and you can save a limited number of maps for individual SNPs to your account.

     Finally, it is important to remember that there are other reasons for the adoption of a surname, and in particular death of father and the adoption of a stepfather’s surname, and illegitimacy, where the child is named for the mother’s surname, being the most common.

                              Grieg(Greig/Grig/Gragg/Gregg)/Gregory

Having previously explored the results for Grieg and Gregory I thought that it would be interesting to see if there was any link between them, always remembering that while those who test are a random sample of the population they include a high proportion of those in America who may in fact be related as a result of genealogical connections which have occurred in America/Canada over the last 400 years.

     Figure 5 shows the complete picture of the two surname groups in order to show those results which lie on the right of the Figure and belong to quite separate genetic groups. Starting from the top, 275997 is Gregory but the haplogroup is J which has its origins in the fertile crescent or further east in west Asia. The larger group with 877623 etc contains the surnames Gregory and Gregg – and these have Viking origins (haplogroup I). 239031 (Gregor) is haplogroup T which also originated in West Asia.  Finally, 212046 is from Haplogroup R1a which spread from eastern Europe but may have originated in India.

Figure 5: Gregg/Gregory all results

Figure 6 is a close-up of the much larger collection of Griegs (Greig/Gregg/Grig/Gragg) most of which can be seen beginning with ‘Gr’ while the Gregories are shown as “GY’. What this shows are two discrete large unrelated groups, one of Greigs (including 29673) and one of Gregories (including 8100). Then, apart from three much smaller groups, [the first in the bottom right which has a collection of Griggs, Graggs and a Gregg (e.g. 372126) and the second near the middle bottom of Gregg and Gragg (including 851482), and third, a quartet of Gregories (including 374579)], almost all the rest are either singles or pairs of results which lie some distance from their common point of origin (the split point where their lines divide). Only one pair looks as though they could have been related, namely, 950624 Griggs and N52932 Gregory but they only share a point of origin and are not otherwise a related pair.

Figure 6: close up of Grieg (all variants) and Gregory results

                                Terminal SNPs

Figure 7 adds the terminal SNPs which have, so far, been tested for these two surnames. To understand the significance of a terminal SNP it is important to understand that, in almost all cases, a SNP is a point on the Y chromosome which has become a fixed point in the genetic tree. Very early on in DNA genealogy the SNP R-M269 was the most recent SNP used to classify individual results for the R haplogroup common in Europe, but that SNP appeared sometime between 4000 and 10,000 years before the present. More recent SNP research and identification has been able to create a family tree of SNPs reaching down through time towards the present. So, a terminal SNP represents the most recent occurrence currently known to research. The company YFull publish a tree where some of the SNPs in sequence have approximate dates assigned – however more recent SNPs are mostly as yet undated.

Figure 7: Figure 6 with tested terminal SNPs added

In Figure 7 the four SNPs which are seen in the Gregg results [top middle] are all connected – some of the SNPs shown occurred earlier than others in time but they all connect together so that the individuals whose results they show are close cousins genetically of each other. Current mapping puts this SNP group in Scotland.  The screenshot in Figure 8, which is taken from the Gregg Block Tree shows how the lines spread out from the common terminal SNP BY135575 (courtesy of familytreedna.com). 

Figure 8: FT108241 as terminal SNP with later terminal SNPs BY212591 and FT106941 in the other individuals

The large Gregory group has a different terminal SNP – there are only two individuals in this group who have tested but their terminal SNPs are likely to proceed from BY34346, a completely different origin from the Griegs (all variants). All the other terminal SNPs shown for Gregories are either unique to an individual or shared by a small group. In some cases, for example 214992 Gragg and Gregg BY173545, there is a close genetic relation but the Gregg individual has acquired a later terminal SNP.

                                  Descendants of Rob Roy MacGregor

In the last two DNA blogs I have shared some of our DNA research to try to identify an individual who is definitely descended from Rob Roy MacGregor, and discussed some of the difficulties in verifying descent through paper genealogies and available records. As well as trying to identify a Rob Roy descendant we are also hoping to identify the Glengyle family of MacGregors of which he was a part.  In my report, I explored the possibility that an individual participant in the DNA Project was descended from Rob Roy’s grandson James who had a son Donald but suggested that it was a concern that this Donald who married Elizabeth Stewart in 1804 was probably too young to be the correct one, and, in addition, terminal SNPs suggested a closer connection to the Chief’s line of Glencarnoch. I am grateful to John Andrew Hutchison who contacted me with a huge amount of information from years of research into the various families who lived in and around Balquhidder and surrounding parishes. John has a family connection to the MacGregors and has been trying to link families together across generations, a task made much more difficult when individual members of families moved around and away from their ancestral homes.There is simply too much information from John to try and summarise here in any meaningful way, particularly as it involves consideration of a large number of different MacGregor families in historical records, as well as family records going from the 18^th century onwards. However the main thrust of the research suggests that Donald who married Elizabeth Stewart was  probably not of the Glengyle branch but possibly of the family who were designated in the Balquhidder children list of 1787 (which I repeat here complete below) as of Brackley, which could explain our participant’s close genetic connection to the Chief’s line – of Glencarnoch  - and/or possibly to the Roro family.    

This list is found in Central Archives Stirling with the call number PD60/651. It is incorrectly dated as possibly 1830.

A list of the young children of the tree of McGregor with their names and Designations within the parish of Balquidder [all listed McGregor unless otherwise stated]

Bracklys family [sic]

Inverlochlairgemore

Patrick

Robert 1

Gregor 2

Donald 3

Alexander 4

William 5

Alexr in Lickseridan

Peter 1

Alexr 2

Dun 3

Alexr in Gartnafuaran

John 1

Duncan 2

James 3

Gregor 4

Malcom 5

Dun in Inverlochlairgemore

Donald 1

Alexr 2

Peter 1

John 2

2 orphans

Dun in Craganmore

James 1

Alexr 2

Peter in Milltown

Dun 1

Alexr 2

Donald in Rinacraig

Donald 1

John 2

Dougal 3

Alexr 4

Alpin 5

all orphans

Dun in Rinacraig    

Duncan 1

in all 27

Dougal Keers Faimily

Duncan in Minachallwaroo

John 1

Robert 2

Patrick in Blarerioch

Malcom 1

Dougal in woodend Stronslany

John 1

Gregor 2

Donald 3

Malcom 4

Gregor in Gartnafuaran

Dougal 1

Gregor 2

Duncan in Stronair

Malcom 1

Alexr 2

Robert 3

Allan 4

John in Criganmore

Gregor 1

Duncan 2

James 3

Twins John and Donald 2

Donld Coock in Ruskachan

Dougal 1

James in Ruskachan

Duncan 1

in all 20

McGregors Faimily

Donald in Ardchullary

Hugh 1

Robert in Adchullary

Donald 1

James 2

Robert in Rinacraig

Peter 1

John 2

Alexr in Rinacraig

Peter 1

John in Rinacriag

Peter [ ]

Robert in Achtowmore

James 1

John in Achtowmore

Hugh 1

Hugh in Achtowmore

John 1

in all 9

Roro Faimily

Duncan in Stronslary

John 1

Donald 2

Peter 3

Hugh 4

Duncan in Balcnoick

Alexr 1

John in Lichscridan

William 1

Alexr in Dalreach

John 1

Hugh 2

Hugh    Drover

James 1

Patrick Decest Balifoil

Duncan 1

John 2

Donald 3

in all 12

A mistake in Brackly Faimily that they not aded togather (sic)

John Decest in Gartnafuaran

Donald 1

John 2

Robert 3

Jobert (sic) 4

Donald in midleachton

Peter 1

Hugh 2

John 3

John Decest in midle achtow

Duncan 1

John 2

Gregor 3

Dun in Glentarkan

James 1

Peter 2

John 3

17

Brackly Faimily

27 (+) 13  (=) 40 of them

As usual I end this blog with the offer to compare a single result with up to 11 others using the spider charts, or to offer further advice on further DNA testing, particularly Y700 and terminal SNPs (and consequently the relationship of differing groups of the same name to each other). Contact me on richardmcgregor1ATyahoo.co.uk (substitute @ for AT).

I am grateful to Central Archives, Stirling, familytreedna.com. and the authors of the spider diagram program Splitstree, Daniel Huson and David Bryant.

Welcome to the MacGregor Project DNA annual blog for 2021.  This past year with the arrival of the virus has had an inevitable effect on the progress of the DNA project as everyone has been concerned to deal with the changes in lifestyle which have been needed.  I thought therefore that I would use some of this update to explore some ideas, opportunities and limitations of autosomal testing – typically called ‘All my Ancestry’ or ‘Family Finder’, and also to share some thinking on the MacGregor Big Y results.

     It is well known now that only males carry the Y chromosome and that this particular part of DNA is usually associated with surname (except for cases of adoption etc). It follows that for surname studies only males with the surname of interest can be tested for connection with a specific surname. For more recent family connections it can often be possible to find a male of the surname to do the test on a person’s behalf but going further back in time does not work so easily when thinking about autosomal (i.e. Family Finder) testing as I will explain. Inevitably genetic scientists will say that the following explanation does not cover all the possibilities and in some senses tends towards simplification: there is a certain inevitability about this when dealing with a wide and diverse audience.

     I have often been asked “why can’t I  be part of the surname project because my great great grandmother was a MacGregor?”, or “my ancestor was Rob Roy, can I do the DNA project to prove it (where the connection is through Rob Roy’s daughter or granddaughter)?”, or “why do the results of Family Finder not show in the surname project?”, or “why are surnames not included in the surname project?”.  I hope to answer all these with the following text.

     Everyone inherits autosomal chromosomes from their male and female ancestors. The proportions received from each ancestor varies, and also these tend to disappear over time as they are replaced by the DNA of succeeding generations. Otherwise, we would have millions of bits of autosomal  chromosome going back into prehistory: having said which it IS possible that some of a person’s autosomal chromosome has survived from prehistory to the present day. Every generation doubles the number of forebears which you have: 2 parents, 4 grandparent, 8 great grandparents, 16 great great grandparents and so on. As you go back in time that number increases, so, around 1300, in the Medieval period, each person will have over a million ancestors and working on the principle that most people are related by ‘6 degrees of separation’ there must be many common ancestors between any two individuals.   

      We all have 46 chromosomes in total and two of these are the X and the Y: if you got a Y and an X chromosome from your parents then you are a male, if you got two X chromosomes then you are female. That leaves 44 chromosomes (2 x 22 pairs) for Family Finder to compare.  Family Tree DNA’s Family Finder distinguishes between matches which arise from autosomal chromosomes (the 22 pairs) and the X chromosome and results are therefore shown for both autosomal chromosomes and also for what FtDNA refer to as X Match

[But note that, surprisingly, of the 5 matches on X Match out of the first 1000 Family Finder matches which I have, there is no detail in the graphics provided – I would be interested to know if others have found this to be the case also). 

To explain this further let’s now take a specific example or two:On the familysearch.org website you can create your own  fan chart of ancestry, and in an article Jessica Grimaud presents a filled-in fan chart: (https://www.familysearch.org/blog/en/genealogy-fan-chart/)

 Fig. 1 Fan chart from www.familysearch.org

There is an amazing amount of genealogical material presented here and it is relatively rare for a researcher to be able to document quite so many generations.  As a  comparison here is my own fan tree which lacks detail from my mother’s side because I have not been able to ascertain with any certainty her Russian German mother’s ancestry.

Fig. 2 Fan chart Richard McGregor

I will begin by discussing an example of how this fan tree was used to identify a common ancestor with myself in Family Finder.  I do not have a fan chart for the person who I match but there is a family tree for this person deposited on the Family Tree DNA website which contains sufficient information.  We are listed as being 3^rd to 5th cousins and these are the places that we have autosomal matches:

Fig. 3 Chromosome matches for myself and the other participant

This may look at lot but when compared with someone who is a 1^st to 2nd cousin you can immediately see the difference. If you count the number of shared chromosomes in Fig 4 and Fig 5 you will see that the latter has many more. That indicates the first problem – the further back in time you go the less SNPs in common you will have with someone else and if it is 5 or 6 generations back then you have to work out which of up to 64 ancestors you actually match to:

Fig. 4: Richard McGregor matching chromosomes for 1^st to 2^nd cousin

Going back to the Fig 4 Family Tree DNA do a version of these results in graphic form which comes out as below (the shading represents the areas of match):

Fig. 5. Graphic form of shared SNPs between Richard McGregor and Family Finder match 3^rd- 5^th cousin

Working back with our researched genealogies we came back to a meeting point with a Georg Konrad Major born in 1797, two of whose children started the separate lines we descend from.  I have over 4000 matches on Family Finder and only a few of these can actually be traced back to a specific individual shared between us – part of the reason for this is that not all participants share ancestral names or genealogical trees on the website.  I have very few 1^st to 3^rd cousins and while it has been relatively easy to find the genealogical connections with these, when the relationship suggested is more distant, it becomes rather more difficult.   There are quite a few examples where Family Finder participants have given ancestral names which match with those in my tree but none has led to a definite link. 

 

Hopefully it is obvious that finding specific links with Family Finder relies on having access to genealogies which are as complete as possible and the closer the genetic connection the easier it is to identify the connection. When we are trying to identify a common ancestor who lived more than 250 years ago,  it is vital to  be able to eliminate other possibilities.

*    *     *

Update on MacGregor DNA

This year has seen more participants upgrading to the Y700 test which has resulted in an expansion of the genetic tree of connections. There have been a large number of terminal SNPs identified, and on the grid at www.familytreedna.com/public/macgregor these are shown in the green colour. Sometimes these just represent a SNP that has been singly tested and is not the terminal SNP, and the best way to determine if you are looking at a terminal SNP is to assume that it will have a haplogroup (such as R or I) followed by a BY, FGC, or FT number [Big Y, Full Genome Corp or FamilyTree]. Quite often individuals who have tested to Big Y have SNPs which are not shared with others who have Big Y tested: these SNPs are known as Private SNPs and will be so identified until there are other participants found who have any of these SNPs – these will then be shown as shared SNPs and will indicate common ancestry between the individuals who bear them. When there are no SNP matches  between individuals the Big Y results default back to the most commonly held SNP which is why some of those genetic MacGregors who have tested with Big Y default back to S690, which is the SNP that all genetic MacGregors have. As more results come in an individual with S690 might  find that one of the Private SNPs becomes the terminal SNP. As I just indicated, in that case it might be reasonably assumed that the two (or more) individuals concerned share a more recent ancestor normally bearing the surname of interest.

Below I have copied the Big Y Block Tree graphic which is currently shown for the MacGregor group descending from Ian Cam. It is based on the format developed by Alex Williamson [The Big Tree]. I have removed individuals’ names for data protection reasons – however the terminal SNP can be cross checked on the results grid link given above to identify the earliest known ancestor where this has been filled in by the individual participant.

 Fig 6: Block Tree (current) for MacGregor Ian Cam group

As another example I have chosen a McGregor result from what is labelled ‘McGregor distant group’ on the results grid:

 Fig 7: Block tree diagram for a McGregor in the McGregor distant group

In this case the surname of participants varies across the grouping , starting with McKellar for the first two lines of descent followed by Robertson; no name given; Campbell; Stewart; McKellar, McKellar (which is the McGregor participant’s shared match); followed by McKellar; then unnamed at the far right,  All Big Y participants can see their results presented in this way through their results page. However, it is important to say that if  an individual’s terminal SNP has been determined without going through Big Y then this block graphic is NOT given in the participant’s results because some of the other SNPs which contribute to the Block Tree  will not have been sequenced. For example, I determined my own terminal SNP BY4303 by deduction and testing, and not through Big Y, so I cannot see who else is close to me.

[Note: McKellar is considered a sept of Clan Campbell and is found across Scotland with authorities variously placing the origin in Argyllshire, Angus and/or Strathclyde].

Finally a McGregor distant result, but with a known Sutherlandshire connection:

Fig 8: Block tree result for a McGregor in the McGregor distant group whose forebear came from Sutherlandshire in the north of Scotland

In this case the rest of the Block Tree group matches are almost entirely called Nicholson (except for a solitary MacDonald).  The individual’s McGregor result box does include an additional named individual whose surname is also McGregor – and in that case this participant and the other McGregor share an ancestor in the more recent past. Who this was might be established by comparing genealogies – the difficulty being that many Sutherland parish records are deficient for pre 19^th century.

[Note: the name Nicholson in Scotland is said to derive from a lawyer who lived in Aberdeen in the 16^th century, but there was also a family associated with the Isle of Skye who petitioned the Lord Lyon for recognition – which they got, on condition they were designated Clan MacNeacall of Skye]

Our own analyses which have been conducted by Prof. Neil McGregor is shown below (as at October 2020, although the previous date of May has been retained). This gives broadly the same result as Fig 7 above and represents the current state of our knowledge:

Fig 9: Ian Cam group Big Y results as presented by Prof. Neil McGregor

We have however identified an issue related to the first chip used by FtDNA.

You will see from the grid given below that the top group lack some results that are available for the lower group. We have contacted FtDNA about this:

 Fig 10: Partial results grid Ian Cam group

You will have noticed that this year I have not used spider diagrams as in previous years. This is partly because comparatively little has changed over the year, but also because I prefer to offer individuals the possibility of seeing a spider diagram for their own result which is designed for them alone. You can see these spider diagrams by looking at previous versions of this blog. If you would like me to generate one for you ask me offline and suggest up to 12 individuals with whom you would wish to be compared by reference to their kit number. You can email me on chairmanATclangregor.org [change AT to @ - I just want to avoid spam!).