HIGHLIGHTS. As our Study approaches the second anniversary of its launch back in December 2005, two other milestones have been passed.

We now have results of 100 participants. For about 20% of these it is still difficult to draw conclusions about their ancestry - some because more than 12 markers are needed. For 80% of the remaining 80%, i.e. about two thirds of all our participants, including several whose surnames have changed, we can say that they had a common ancestor who came from the Scottish Borders. But about 15% of participants appear to be unrelated. A likely explanation for some of these unrelated participants, and for our genetically related participants with a different surname, is that they are descended from sons who took the names of their step-fathers.

Meanwhile our host company, FTDNA, now has over 100,000 y-DNA test results, easily the largest y-DNA data pool, and have refurbished their website, enabling the results and analyses of our Study to be sent by bulk e-mail direct to participants.

INTRODUCTION. DNA testing is an increasingly popular tool with genealogists. The test itself is simply a saliva sample from inside the participant's cheek. The privacy of each participant is protected by limiting analyses to the “ancestral line” rather than to the full personal DNA “fingerprint”. Participants’ e-mail addresses are not disclosed to third parties.

The interpretation of DNA test results depends on the transmission of DNA remaining unchanged from generation to generation, apart from small and occasional changes (“mutations”) in one or more of the “markers” that make up the genetic elements of the DNA profile of each individual. The DNA profiles of individual participants can be compared to establish the likelihood of common ancestry, but DNA test results are never 100% conclusive. Confidence in analyses of the results increases as more individuals participate and the testing and analysis techniques continue to improve.

One application of DNA testing is surname studies, in which DNA tests of y-chromosomes from male participants are used to complement and augment conventional genealogical research. Other applications of DNA tests include “deep ancestry” (i.e. identifying descent from Celts, Vikings etc.), and descent via the female line (“mt-DNA”). Neither of these applications is within the scope of this Study.

Since 2000 many Scottish Clans have launched surname studies and associated websites. Such studies offer opportunities for:

• enabling DNA tests at group discount rates with directly comparable results;
• helping genealogists whose conventional research has met a “brick wall”;
• ascertaining the probability of two participants being related;
• identifying branches within a Clan;
• exploring and developing new ideas.

This webpage does not attempt to describe in detail the principles and terminology of DNA tests for genealogists. For further guidance see:

• DNA and Family History
• Family Tree DNA.com
• DNA Heritage
• MUMMA Surname DNA Project
• Graves/Greaves DNA study
• BLAIR DNA Project
• MacGregor DNA Project
• Reiver Families

GOALS OF THE CLAN IRWIN SURNAME DNA STUDY.

1. To facilitate and co-ordinate cost-effective y-chromosome DNA ancestry tests for all those bearing surnames sounding like Irwin, Irvine or Irving, and individuals related thereto.
2. To combine the results of these tests with available genealogical data.
3. To identify DNA profiles of the Bonshaw and Drum branches and sub-branches thereof.
4. To promote, co-ordinate and publish analysis and discussion of the results through the Clan Irwin Association’s website and its quarterly journal, the Holly Leaf Chronicle.

JOINING THE STUDY. Participation is invited from all adult males who have surnames sounding like Irwin, Irvine, or Irving. As surname studies only address y-chromosome test results, females can only participate in this Study if they can persuade an adult grandfather, father, brother, uncle, nephew or male cousin to have a DNA sample tested on their behalf.

This Study has opted to use Family Tree DNA (“FTDNA”) as the testing company because, in the absence of any international standard, it has the largest data pool. FTDNA was established in Houston, Texas in 2000 and now has conducted over 100,000 y-DNA tests, many more than any other company. It hosts over 4000 registered surname studies and provides useful support.

For prospective participants able to trace their ancestry earlier than the 19th century we recommend FTDNA’s 37-marker test at $189. Prospective participants unable to trace their early male ancestry may opt for their 12-marker test for $99; if this initial test indicates a possible relationship we would then recommend upgrading to the 37-marker test for an additional $99 (FTDNA retain their samples indefinitely). Individuals' costs may be reduced by persuading other members of their family to share the test fee. FTDNA also do 25-marker and 67-marker tests, but we do not normally recommend these for this Study.

To participate in the Clan Irwin Surname DNA Study at discounted Group rates, go to www.familytreedna.com> “Irwin” >"Search " > "Irwin Clan", and complete the Group Application Form.

When FTDNA receives your application they will send you a sample test kit and simple instructions on its use. On receipt of your two samples they will undertake the relevant analysis and enter the results into their confidential genetic database, compare these results with those of other participants in this database, and advise you of any exact or close matches. If you complete FTDNA’s Election Form you gain access to e-mail addresses of these participants, and to quantified probabilities of how closely you are related to them. Uploading to “Y-search” posts your results on the web and enables non-FTDNA participants to contact you without your e-mail address being disclosed. One participant in the Study serves as Administrator, seeking to identify family branches, to co-ordinate research, and to promulgate results and associated issues without compromising confidentialities.

STUDY PARTICIPANTS. The Study now includes participants from the Erwin, Genographic and Irish Heritage projects, and participants with other surnames who appear to be close genetic relatives. The first 100 participants’ dates of DNA test and country of residence are:

Participants have volunteered the name, date of birth and place of origin of their earliest confirmed paternal ancestor:

The “Other” spellings are participants’ surnames unlike Irwin but whose DNA indicates their ancestors were genetically related to Border Irvings.

INTERPRETING DNA TEST RESULTS. Prospective participants are cautioned that some DNA test results prove disappointing. This can occur if there has been a “Non-Paternal Event” in the male ancestral line (e.g. adoption, illegitimacy or name-change, which were not uncommon), if the associated family tree is incorrect (yes, it does happen), or if the name was assumed before surnames became strictly hereditary.

The DNA test result of a single participant is of no value until compared with that of another participant so that the probability of the two participants sharing a common ancestor can be assessed. This probability is dependent on many variables, the most important of which are the number of markers tested, the number and magnitude of the mismatching markers, and the number of generations since the most recent common ancestor (MRCA). The following “rule of thumb” is a qualitative indication of the MRCA within the last 24 generations:

“Distance” is the sum of the magnitudes of the mismatches of individual marker values, e.g. a distance of 3 may include three 1-step mismatches, or one 2-step mismatch plus one 1-step mismatch. Special rules apply for DIS markers 385, 464 and YCA.

In fact the probability of two participants sharing a common ancestor is a complex mathematical function that also depends on whether the rates of “mutation” of the mismatching markers are fast or slow - faster mutating markers being indicative of a relatively recent MRCA. When the likely number of generations since MRCA is already known from genealogical research, FTDNA’s “TiP” algorithm takes all these factors to account and calculates a quantitative probability of any two participants being related.

When comparing more than two results it is convenient to select the DNA profiles (“haplotypes”) thought most likely to represent different branches of a family. This selection is undertaken by the Group Administrator.

As the number of markers analysed is increased from 12 to 37 to 67, so the TiP probabilities of common ancestry of two participants tend to polarise: thus TiPs of 10% or 90% with 12 markers may move to 5% or 95% with 37 markers, and to 2% or 98% with 67 markers. But there are many exceptions to this generalization, and the movement of 12–marker TiPs nearer 50% are even less predictable as the number of markers analysed is increased.

When interpreting their results participants should bear in mind the possibilities of adoption, illegitimacy, name change and genealogical error. Confidence in the results will increase as the number of other participants continues to grow.

TEST RESULTS. The first 100 y-DNA test results of our Study are listed in Table 1:

ANALYSIS OF TEST RESULTS. The same results are listed in Table 2 in six groups, and within each group in order of probability of relationship with the modal haplotype `BA’ (the most typical result). Columns (1) & (2) identify the haplotypes and groups; columns (3) & (4) list the personal identifiers assigned to preserve the anonymity of each participant; columns (5) - (9) summarise details that each participant has provided to the Group Administrator about their Earliest Confirmed Paternal Ancestor; column (10) shows the number of markers which each participant has asked to be analysed; columns (11) - (14) show “distances” from the modal haplotype `BA’; columns (15) - (21) show various probabilities calculated by FTDNATiP of each participant having common ancestry with indicated haplotypes; columns (22) & (23) list close matches with other participants.

Analysis to 37 markers has proved the most popular and cost effective:

12 markers (for most participants) and 25 markers (for some) have proved too few, while alas the benefits of going to 67 markers have proved disappointing. Only 15 participants have managed to post their pedigrees on to the FTDNA website.

In recent months those participants whose DNA test results imply a close genetic relationship (0/37 or 1/37) have contacted each other to seek genealogical relationships. So far these initiatives have only identified one pair of 5th cousins (at 1/37) and one pair of 6th cousins (at 2/67). In contrast our Study includes three pairs of brothers whose genetic relationships turned out to be 0/25, 1/37 and 2/37. These examples show clearly the difficulties in combining genetics and genealogy.

Despite these disappointments, by end October 2007 it has been possible to identify the DNA profile of three branches of the name (Borders, Orkney and Munster). The first 100 y-DNA results in our Study can be characterized thus:

Within the Borders Group`BA’ has been assigned as the “modal haplotype”, i.e. the digital profile of the most common count for each marker in the Study. This may well be the profile of the common ancestor. `BB’ is a senior representative of the Bonshaw branch; and `BD’ is a senior representative of the Dumfries branch, the current owners of Bonshaw Tower and allegedly the 1oth cousin of ‘BB’. But as ‘Bd’ and ‘BD’ are full brothers with genetic distances of 3/37 and 5/37 respectively it is difficult to say whether DNA has either proved or disproved the traditional genealogical relationship. Participants in this group are listed in order of probable relationship with `BA’, but this does not infer that `BA’ is “senior” to `BB’.

Several participants with other surnames are included in this group as they probably share a common ancestor: Armstrong and Elliot are common Borders surnames, while Byers was a common name in Annan. Ancestors of these participants may have had Irving step-fathers or been adopted or illegitimate. The Cahill and Hamblen participants know their relevant recent ancestry included such events. Errand may be a name change.

Our Study has not yet attracted a participant with confirmed Drum ancestry. Several participants in the Borders group have unconfirmed traditions of Drum ancestry, but no distinguishing markers have yet been identified.

Three participants claim descent from Orkney ancestors. One, ‘OP’, has a confirmed genealogy back to 1460, but it is still unclear whether he is related to ‘BA’; a second is 4/37 from with ‘OP’; but a third is 17/37 from ‘OP’ and so must be considered unrelated.

Our Munster Group has two participants with Co. Limerick ancestry. One, ‘IM’, was named O’Ciarmhachain. It is now clear that this Group did not have Scottish origins.

On present evidence the fourth group of participants have to be considered as “Unrelated” to the above groups, or other participants within this Group (apart from the two cousins). Some of these participants may turn out to represent “new” branches, while others may have acquired the name Irwin by some “Non-Paternal Event” (see above).

The genetic relationships of the remaining two groups are Unclear – in the case of the fifth group because 12 markers is too few to allocate them to one of the first four groups, in the case of the final group because 25 or even 37 markers has not proved sufficient to demonstrate a clear genetic relationship to ‘BA’, ‘OP’, ‘IM’ or some further group yet to emerge.

Not included in these tables are the many individuals whose surname does not sound like Irwin but who have “close” y-DNA matches with participants with less than 37 markers, or have 37-marker results with distances greater than 4. Such “matches” can probably be attributed to random mutations and have no genealogical significance.

CONCLUSIONS. Several points are emerging from this Study:

1. The involvement of 100 participants within two years of the launch of our Study is testimony to the interest in using DNA tests to further our understanding of the genealogies of our surname.
2. 37-marker analyses are usually the most cost-effective. 12 and 25 markers are insufficient for most participants; 67 markers have yielded little additional information.
3. As expected, the spelling of our surnames today is not a reliable indicator of the branch from which one is descended, at least in USA. (NB Within Scotland Irving is still more common on the Borders, and Irvine elsewhere.)
4. For about 10% of our participants it is still difficult to draw conclusions because 12 markers is too few, and for a further 10% because of other reasons.
5. 80% of the remaining 80% (i.e. about two thirds of all participants) are descended from a common ancestor from the Scottish Borders. This proportion having a common ancestor is higher than for many other Scottish Clans.
6. However about 15% of our participants appear to be unrelated to any branch or, apart from two 2nd cousins, from each other. Several reasons may account for this, including some ancestors being named after their stepfathers.
7. It is proving more difficult than anticipated to identify separate branches, and to identify genealogical relationships of participants with close genetic relationships (0/37 or 1/37).

Despite some of these findings being more disappointing than many participants had hoped, I remain optimistic that our Study can continue to make interesting progress:

• Although the initial surge of participants following the launch of our Study has passed, a steady flow of new participants continues. At the time of writing a further ten test results are anticipated.
• As more results accrue, further trends are likely to emerge, genealogical relationships identified, and uncertainties resolved.
• Several branches of the family with long genealogical pedigrees have yet to be attracted to contribute their DNA samples to our Study.
• We may also be able to identify a further Irish branch without Scottish origins and a German branch, and attract an Urwin from Northumberland with Borders ancestry.
• The science of analysing DNA test results is still evolving.

ADMINISTRATION. This study is administered by James Irvine who is responsible for liaison with participants and with FTDNA, and for the promulgation of results and analyses through bulk e-mails to participants, the Holly Leaf Chronicle to CIA website, and the CIA website for all.

All queries should be addressed to:

Participants are encouraged to post their pedigrees for other participants to view. Instructions for uploading GEDCOM files are given under the second orange point on the Personal FTDNA page. Some Family Tree Maker files can be exported to GEDCOM thus: under “File” select “Copy/Export”, give the file a name, and change the “Save as type” to “GEDCOM (*.GED)”.

Note that The Clan Irwin Surname DNA Study is an independent study undertaken on a zero cost basis. The Clan Irwin Association endorses its goals but not necessarily the views expressed on this webpage. Furthermore, neither the Association nor the Administrators of this Study, past or present, claim expertise in genetic studies; they have no involvement with FTDNA nor have they independently verified the claims of this organization, and they take no responsibility for any results, analyses or disclosures of private information that may arise directly or indirectly from this Study. Participation in this Study is welcome but strictly voluntary, and each participant remains responsible for satisfying themselves on issues of reliability and confidentiality, as neither the Association nor the Administrators undertake any representations or assurances thereon.