Armstrong DNA Update 18 March 05

It's time for an update regarding the Armstrong and Border Reivers Y-DNA Project. We have a new set of Armstrong Y-DNA Results for Kit #29377.   I'm going to take this opportunity to draft a "primer" for participants in the project.  Warning: this message is getting rather long...  There may be some problems sending this message through the various servers and virus filters.   If it doesn't make it though to you, I will be posting this Primer as a General Comment webpage very soon.  

As many of you know, our "A" List Administrator, John D. Armstrong, has taken on the job of DNA Project Administrator.    Bob Armstrong and I are co-administrators.   I am also co-admin for the Elliott and Border Reivers DNA Project, administered by Jim Elliott.    I am also the Project Administrator for the Strong/Strange Surnames DNA  Project and for the Donegal Bay Geographic DNA Project.   Jim Elliott is also involved as a co-administrator for the Liddel DNA Project.    The interlocking nature of our projects makes it possible for us to share information and DNA data results quite well.    

Jim Elliott has accumulated about 1000 (that's one THOUSAND !!) Y-DNA test results in his study of the Border Reivers.  You can see the results of his study at:

http://freepages.genealogy.rootsweb.com/~gallgaedhil/elliott_border_reivers_dna.htm

and the linked pages.    The Border Reiver's DNA Project has demonstrated that there are a wide variety of Y-DNA Haplogroups to be found in the Anglo-Scottish Borders area.   The largest majority of the Y-DNA Haplotypes belong to Haplogroup R1b, which seems to be Celtic in origin.   Another large proportion of the DNA Haplotypes belong to Haplogroups R1a, I, I1a, I1b, and I1c.  These Haplogroups seem to be associated with the various Viking and other northern European groups which invaded the British Isles through past centuries.   And, finally, there are certain Haplogroups,  and "sub-clades", including C, E, G, J2, and R1b1c, etc., which seem to have either arrived in the British Isles with Roman legionary troops from various parts of the Roman Empire spread around the Mediterranean Sea, or to have arrived with some of the barbarian invaders or the Normans during the 1st millenium, AD.

What does all that have to do with Armstrong Y-DNA?   Well, it helps us identify certain DNA test results with the various ethnic origins... giving some insight into the origins of the lineages...  and a means of identifying particular sets of results with particular lineages.  That also has consequences for the traditional paper researcher, in that identification with particular haplogroups and lineages helps eliminate others, and increases focus for further research efforts.   Close matches can also point to common ancestry with other test subjects.    So, what have we got to this point in time?

The Armstrong Results are posted on my webpage at:

http://freepages.genealogy.rootsweb.com/~donegalstrongs/armstrong.htm

These results are drawn from several sources, including members of the Strong DNA Project, the Elliott and Border Reivers DNA Project, and the more recently established Armstrong DNA Project.    To this point in time, we have eleven sets of Armstrong results.   Additionally, there are at least two more currently in progress.   Unfortunately, one participant died before a sample could be obtained; the researcher involved is attempting to obtain  substitute participation by the decedent's son.    Referring to the eleven sets of results to date, we have representatives of four Haplogroups.... I, R1b, R1b1c, and, subject to further testing, either a variant of I or J2.

The two definite "I" kits most probably descend from Vikings.   The two R1b1c and J2 kits probably descend from either the Roman Legionaries who were stationed in or retired in the British Isles; or from such legionaries who were associated with the predecessors of the Normans while still in France .    The largest group, seven R1b kits, probably represent the original Celtic stock which was present in the British Isles  before Roman times.      It seems particularly remarkable that the Armstrong R1b kits actually vary very little from each other... and have certain characteristics which so far distinguish them from other R1b haplotypes associated with other Border surnames.   This is in sharp contrast to what might have been expected of folks descended from Reivers...   so much hype about rapine, trial marriages, fostering, adoptions, shifting allegiances, and the like may be over-blown.   We need a lot more evidence before we can say that for sure, but please take a look at my Border Reivers DNA Results page:

http://freepages.genealogy.rootsweb.com/~donegalstrongs/reiver_families_r1b.htm

and compare certain features of the various surnamed results appearing there.   All of the results on that page are R1b results.   We have a wide variety of haplotypes there, but we have identified certain "dominant" or "prevailing" haplotypes as "Assumed" general haplotypes associated with certain of the surnames.   See the two main Elliott haplotypes, the Irvine haplotype(s), the Liddel/Lyttle/Little haplotypes, and the Armstrong haplotypes.

Most of these R1b results are quite similar in that the numbers ("alleles") for particular DNA "markers" are the same [see for example, DYS385a, with an allele of 11 for the most part; and see DYS388=12, with a single exception].  However, I have highlited certain key differences either with a light yellow background, or with red type.  Notice  the Armstrong DYS449=32 results, with one exceptional 31.  That "32" distinguishes the Armstrongs from all of the other surnames.  Similarly, see the two Elliott Haplotypes, where DYS458=18 in one haplotype, and 17 in the other.   There are two emerging prevailing Irvine Haplotypes, differing at DYS391, where there are values of 10 vs11.   The Liddels are distinguished by DYS385b=11 and DYS389-2=30.   There are other emerging haplotypes as well.... but this is more than enough to illustrate the point.   For the most part, the DNA results associated with these haplotypes tend to run true to their prevailing "Assumed" Haplotype.  I have imputed the various "Assumed" Haplotypes from the prevalence of marker values involved... looking for the modal allele values occuring most often in connection with the surname in question.  There is a mixture of observation, science, and art involved... enhanced by a bit of experience.  {:-)  And, as the database grows, changes DO sometimes occur.

Our latest set of Armstrong Results, for Kit#29377, fits right in with the Armstrong "Assumed" R1b  Haplotype.   Take a look at the kit on the Armstrong results page:

http://freepages.genealogy.rootsweb.com/~donegalstrongs/armstrong.htm

Notice on the first set of markers, Kit#29377 varies not at all from the "Assumed" Haplotype.  On the second set of markers, he is off just 1 step from the "Assumed" Haplotype.   An excellent match, not only with the imputed "Assumed" Haplotype, but with the other Armstrong R1b kits, which are similarly close to the modal Armstrong R1b values.    What can we do with this from a genealogical viewpoint?

First, we can eliminate the necessity of considering relationships with the four of the eleven sets of results which are not R1b and don't match... the I, J2, and R1b1c kits.   [Note, given enough other participating researchers, we will have other research focus available for those kits... they concentrate on the I, J2, and/or R1b1c kits....  and this can lead to some interesting possibilities.... other matches within the Armstrong surname, or matches with other surnames.]

Second, we can share contact information and paper research with the other researchers involved in the Armstrong R1b kits.   There are various ways to do this... direct contact via the "A" List or private emails or snail correspondence.   Participants can ask the project administrators which other researchers are involved in particular kits.   A good step is to Upload the kit results to Y-Search.  This can be done by going to the participant's Personal Page on the FTDNA website, clicking on  the Y-DNA Matches tab, and you then see an
explanation and a link for the upload to Y-Search.  By following the instructions, you can automatically upload the results to a new page at Y-Search;   you will be given some options,  such as opportunity to provide geographic migration information, pedigree information, and three choices regarding contact with you: No Contact; Direct Contact with a publicly displayed email address; or the most popular choice, Contact only through FTDNA, which can forward messages to you without disclosing your email address.     It is also possible to go to the Y-Search main page, http://www.ysearch.org/ and then clicking on the "Create a New User" tab.   You will be taken to a page where you can directly upload each marker and enter the personal information mentioned above.  Note, there is another database, called Y-Base, run by a competing company, Heritage DNA, to which you can also upload the results by going to the following page, and entering results for each marker along with certain other information.  See: http://ybase.org/default.asp

Third, we can try to match lineages.  This is the importance of having good pedigree information available for each kit.   It is the stuff of traditional genealogy.... where is the Common Ancestor in two or more lineages?    Can we link onto other longer pedigrees, extending further back in time?   [And notice... we can verify the validity of the research... someone may claim a connection... but DNA testing can show his line to be unrelated!]

Fourth, in the absence of a clear identification of a Common Ancestor, we can use some of the tools made available to us by FTDNA to try to estimate the time period in which a common ancestor may have lived.   [Obviously, this does not identify the Common Ancestor... we still have to do paper research for that.   However, we can focus on a most likely time period, and hopefully, given sufficient migrational and other background information, identify a particular geographic locale in which to do further research.    Again, sharing of such information between researchers is most helpful.]    What tools are available from FTDNA to help estimate the time to a Common Ancestor?  FTDNA has developed a Family Tree DNA Time Predictor calculator program, FTDNATip, which can be used to estimate the probability a participant shares a Common Ancestor with another participant, within differing perameters including centuries elapsed and # of known generations in which a Common Ancestor did NOT occur.  Built into the calculator program is a factor based on direct research of the mutation rates of the particular DNA markers being compared.   Below is an example of the use of the FTDNATiP calculator; Note, for the second calculation below, I ESTIMATED that there was no Common Ancestor known between the two kits used as an example during the previous 4 generations.   This estimate can be changed based on actual knowledge: