Where are we now? Update #5

This update concentrates exclusively on the 37 marker test results. The purpose is to show what we think we have learned so far about the genealogical connections which have been suggested by the results. The problem with mutations, as I have said elsewhere, is that they are, by and large, random but seem to stay constant within families. So if individuals with the same surname compare results it should be possible to observe groupings, particularly if the surname group comes from a small geographical area. The table of all the 37 results is given below.


Table 1 - 37 Marker Comparison (click to enlarge)

In this order, as given by Family Tree DNA in the table of results (at www.familytreedna.com/MacGregor/public), those who are closest to each other in genetic terms are grouped together – most of the time. However, if there is a more unusual mutation, such as on the second marker where 24 for example becomes 23, then that person will appear to be quite distant from the larger group, since the 23 conditions position in the table. This is what has happened to kit 24029 in the MacGregor group and also 17041 in the Orr group. Both these individuals are clearly attached to their main surname groups as will be seen in the next charts and diagrams.

In this update I have included, for the first time, an indication of the earliest known ancestor for each of the group members and this will be found at the end of this report. I am satisfied that this does not compromise anyone’s privacy since it is not possible to work out who each testee is unless he (or she) chooses to make him/herself known.

The next table (Table 2) has been generated by an excellent online program created by Dean McGee with whom I have had correspondence, because there is a supposed link between McGehee and MacGregor (as yet unproved by DNA). This program uses the data provided by Family Tree DNA on mutation rates and calculates the time, in years, to the common ancestor. In this table I have only included DNA results which might be directly related by common MacGregor ancestor. The line in bold (on kit 2124) is the Glencarnoch MacGregor line (as of the present chiefs). I have used a 75% probability of relationship based on a generation length of 33 years. These are rather larger figures than other groups use but I have based it on known timeframes. So for example we know that the Stirlings [kit 9290] could have split off from the main line no later than 1680, and the Glengyle branch must be before c1500 [kit 1774].


Table 2 - McGehee/MacGregor Relationship (click to enlarge)

These figures should not be taken as implying definite dates since, as I have said often, mutations happen unpredictably and possibly according to changes in circumstances in ways we don’t fully understand yet. It has been suggested for example that the DNA of emigrant families is likely to show more mutations than those who remained in the home country. I have no idea if there is any scientific basis to this statement at all but if it IS correct, then the time frame to nearest common ancestor would be shorter.

These charts were constructed using the computer program created by Fluxus Engineering. As I often say we must remember that is just one interpretation of the data by a computer program so, as always, the results can be interpreted in different ways. To counteract this I also give some other similar charts which suggest some slightly different interpretations of some of the data. By the way, the 'mv' points mentioned in the following discussion are essentially points where the DNA divides to go different ways. They can be thought of as points where two brothers diverged, although this is not strictly accurate (but possible).

Chart 1 is 'rooted' in the ancestral DNA sequence associated with the hunter gatherers moving north and west to the Atlantic fringes, following the retreating ice – Spain, Basque country, Brittany, Ireland and Great Britain. This is called the Atlantic Modal Haplotype and is the most common DNA grouping in those countries. In the chart it is called ATLANT. Over time, two markers in the MacGregor group changed from having 11 repeats in the DNA sequence to having 10. These were marker 391 and 385a. On Chart 1 below, this point is labelled ‘Point of Common 10 10 Ancestor’. Individuals to the left of this are almost certainly descended from John Cam d..1390, son of Gregor, name father of the clan, or from his children.


Chart 1 - Atlantic Modal Haplotype (click to enlarge)

The grouping of the entries on the chart suggests that there was a split of ancestral lines at the point labelled mv41. Since we know that kit 2124 is the line which includes Glencarnoch, and kit 1774 is Glengyle, this is consistent with a very early ancestral split, as the genealogies suggest. We can be fairly confident that kit 20123 is a Glengyle. The earliest known ancestor of 20123 lived in Kilmallie parish, Argyllshire, where there is now the intriguing possibility that the story of a grandson of Rob Roy being settled there is therefore correct, if this is the same family.
The next split occurs with kit 2125 and the testee here believes, but cannot prove, Roro family descent. IF this tree is correct and if 2125 is Roro then it would follow that all others after this would also be Roro. This is a very attractive idea but as I will show depends on 2125 being where it is, and this might not be correct.

There may be a problem with the position of kit 4714 if their family belief in Roro descent is true. However since neither 2125 nor 4714 know for sure we need a testee to come forward who has definite documentary evidence for Roro connection in order to help verify this.

There can be no doubt that kits 2909, 10897, and 16858 are either related to the Glencarnoch line directly ,or, share a more recent common ancestor with that line. Again a direct connection between kit 3346 and Glencarnoch is suggested by this chart. However, as you will see, other charts make this a slightly more distant relationship. Nevertheless a connection between the Macgregor line found in Weem parish in the late 18th century and the Stirlings who found their way to Dunblane by at least the 1670s, appears possible.

An intriguing possibility is suggested at mv11 - that there was an earlier Gregor split which gave some Macgregors and some Greigs their connection, perhaps in the the 11th or 12th centuries. Then the split at mv5 could be where various clans have their joint kingly origins. This is very much speculation, BUT, the genetics does indicate common ancestries between clans - but the linking dates are not clear.


Chart 2 - Atlantic Modal Haplotype: Another View (click to enlarge)

So much for Chart 1. Chart 2 looks a little different. The Glengyle branch is omitted as well as all more distant MacGregor lines. The joining of tree lines in this chart suggests that there might have been an older split before that of Glencarnoch, and that the oldest MacGregor line here is preserved in the DNA of kit 4714 , with another older split at mv2 leading to kits 10897, 3346, the Stirlings (9290 + 13621), and 24029. As I said earlier, it's not possible yet to say for sure which interpretation is correct.

Another significant difference in Chart 2 lies with 2125 which here comes out as a branch line, not a main trunk. Nevertheless if 2125 is Roro then so must all others after mv10 be also, except for possibly 13429 and 18760. It then follows from this chart, that Glengyle is a Roro offshoot. Clan historians think this is rather unlikely. Charts 1 and 2 imply that BOTH 2125 AND 4714 cannot be Roro as the genetic distance is too great.

Charts 1 and 2 were prepared by Neil McGregor a project member in Australia. Chart 3 has been prepared by me. I have included a number of other 37 results for comparison.
Chart 3 is an 'unrooted' (see note) chart containing only group results. Again notice the nearness of the supposed Roro 2125 to Glengyle 1774.


Chart 3 - Unrooted Group Results (click to enlarge)

An intriguing link, which is probably a 'red herring,' is the line connecting Glengyle, whose patronymic was McCoulkeir, that is 'son of Dougal ciar' and our Macdougall testee 21971. Apart from this the same basic groups are maintained as in Neil's charts. The earlier Macgregor split - above 4714 - is doubled in size (that is, includes twice as many people, but time distances, roughly represented by length of line, are rather longer.

With the results currently available we can probably group together the following results: 3346 with the Stirling results - these lines may connect with Glencarnoch but we cannot be certain yet; 1774 and 20123 are almost certainly Glengyle; 2124, 2909, and 10897 appear to be Glencarnock - either that or 2909 and 10897 have very stable DNA! Roro is hardest to explain. If 2125 IS Roro then so should 18760, 22187, 16798, 12596, and 22187 be, BUT if 4714 is Roro then 3346, the Stirlings, and 24029 are this group.

A possibility which can also be considered is that one of the proposed Roro groups is in fact a descendant group from the once numerous Gregor MacIan group - hereditary keepers of Kilchurn Castle under the Campbells in the 16th century.

As I said, we need more testees with a knowledge of their MacGregor branch connection.
Note for Charts 2 and 3: a rooted chart is connected to a known external set of results, whereas unrooted uses ONLY members of a project group.


Chart 4 - Genetic Distance (click to enlarge)

The final chart (Chart 4), which looks very much like a traditional tree, shows the distance ONLY between individuals on the basis of the number of DNA mutations which separate them. It therefore shouldn't be read exactly as a 'family tree'.

What it shows is how near individuals are to each other in genetic time. Note that it suggests a closeness between 4151, 20630, and 5276 which is not correct. Otherwise it is fine except 2125 and 22185 are not quite as genetically close as it suggests. Compare the groupings with Charts 2 and 3 and there are similarities.