O’Shea DNA Project Results
Abbreviations:
FTDNA: The company FamilytreeDna based in Houston, Texas who
organise the yDNA tests for our project (www.ftdna.com/about.html)
Y12, Y25
& Y37: 12,
25 & 37 marker
Y- chromosome DNA tests available from FTDNA.
O’Shea: Includes
all variant spellings such as Shea, Seay, Shay, Shee etc.
Highlighted Text: Indicates
where the text has changed from the previous article
(Rev. 6 of August 2007).
1 DISPLAYING
RESULTS
Results to date (31st Nov. 2007)
are summarised in the separate Table 1. All participants will have received their
results from FTDNA
but although they provide a lot of useful data relating to individual results,
they do not attempt to segregate results into groups. The project
administrators review all results and produce the table.
At the top is seen a row of figures
with the heading ‘Ancestral Modal.’ These figures are divided into groups of 12,
13 & 12 marker
boxes corresponding with Y12, Y25 (12+13) and Y37 (12+13+12) tests. The figures
(known as alleles)
within each box are derived from all results. For instance to decide on the
value for marker 1,
the values for marker 1 of each relevant result directly underneath are examined. This shows
that of the forty five
results, forty four
have a value of 13 and one is 12. Thus 13 is inserted in the Ancestral Modal. (The one different value has been
checked with FTDNA
who confirm it is not an error). The process is continued for the 36 remaining markers. Thus for each marker the Ancestral Modal represents the most frequently
occurring value of all relevant test results. In the case of marker 1 it is quite clear what the result
should be, but this may not be true for other markers where there could be more
than one common value. In such cases all values are given. This doesn’t apply at the
present time although marker 27, currently listed as 11, actually fluctuates
between 11 & 10 as new results are received. The string of figures
relating to each participant’s results is called a Haplotype. In relation to the Ancestral
Modal the 37
figures shown can be considered to be the Haplotype of the unidentified hypothetical
common ancestor of all O’Sheas. The Ancestral Modal is recalculated every time a new
result is processed and thus the figures presented may change with new issues
of this article. As we now
have a reasonable number of results, the modal is probably at this stage pretty
representative, but it will be interesting to see if it will change as more
results are processed.
The next row in the table starts with a column ‘Participant’ on the left. The ‘Code #’ is the kit number assigned by FTDNA when the participant originally applied to join our surname project. Each individual should recognise their own code and set of results, but if you don’t know your code contact us. Participants are also identified by their surname. When produced in electronic form an associated comment box shows the participant’s current base and if known, some genealogical detail. (These notes, if requested, will also be provided in hard copy!) Christian names and precise contact information are withheld to preserve anonymity, but we can arrange to put people in contact with each other if requested.
The next 37 columns contain the headings under which participant’s results are recorded. On the top are marker numbers 1 to 37 and underneath are associated ‘DYS’ numbers. Fourteen of the thirty-seven markers are shaded in bright green such as marker 5 (DYS 385a). FTDNA inform us that these are volatile, mutate more quickly and mismatches in these between individuals are less significant than mismatches in the other markers.
On the extreme right shaded in brown
is a column ‘Haplogroup,’ a term used by genealogists to describe distinct, large population
groups. Results to date in our project indicate we belong to the R1b
haplogroup, the most common in Western Europe. Commentators suggest that all
variants of R1b entered Europe from the east as pre-historic hunter-gatherers,
and migrated along rivers and coastlines and across the ridgeways of high
ground, eventually reaching the Baltic, North Sea, Mediterranean and Atlantic
coastlines, perhaps some twelve thousand years ago. Some participants, such as
#8927, #34730 and #34950, have undertaken specialist haplogroup tests and as expected the result
showed up as R1b. All others have been predicted to be R1b.
2 CRITERIA
FOR FORMING GROUPS & PRESENTATION OF RESULTS:
Anyone joining the O’Shea DNA project hopes that their results will closely match others. Such a match in most cases may be the sole evidence linking the participant to others of the same surname beyond about four to five generations, when the paper trail both within Ireland and overseas tends to hit a ‘brick wall’. For people based outside Ireland it may throw some light on where their ancestors lived prior to emigration. For people resident in Ireland it may indicate how their forefathers migrated from what is assumed to be the original Kerry base. It is not that easy to decide on how results should be grouped as can be confirmed with administrators of other surname projects. For the moment we have decided on the following general criteria, although a certain amount of flexibility is allowed.
· As the majority of participants have now taken the Y37 tests, groups are based on these. Results matching 33/37 (i.e. four or less mismatches) or having an 84% probability of having a common ancestor within 20 generations (500 years) are grouped together. (This information is available from FTDNA). All in a group must match within these criteria. Results matching some but not all in a group are placed in a ‘Subset’ of a main group with a note as to why they have been so allocated. All within such a subset must match 34/37 or have a 95% probability within 20 generations. Results showing a distant link to a main group such as matching 30/37 are placed in a category called ‘Associated’, and are shown underneath the main group and associated subsets. It should be noted that inclusion in such a category denotes an extremely tenuous link.
· Results that don’t match any other are placed in a separate ‘unassigned’ grouping and a note is attached giving an explanation why. Such results may well join new groups as additional results become available
· For those who have done only the Y25 test to be included in a group, results must match all in the group 22/25 (i.e. three or less mismatches). Results not matching all will be placed either in a subset or the unassigned group. Such placement within a group may change if the participant upgrades to a Y37.
· Results for Y12 tests cannot really be compared to Y25 or Y37 and such results are placed in a separate Y12 grouping pending their upgrading.
· Those of different Haplogroups are segregated. At present the majority are in Haplogroup ‘R1b’ with three in ‘I1a’ and one in ‘G’.
· If a result fits into more than one group, the closest match is selected and a note confirming this is inserted.
We have an open mind on how groups are formed
and will review our criteria as more results become available, and as the
science of ‘Genealogy by Genetics’ becomes more sophisticated. The
comprehensive work being done by other surname projects on group formation is
also kept under review. We will consider adopting any standard methodology that
emerges in the future.
Within a group a decision has to be reached as to who should be shown at the top and how the remainder should be sorted. Y37 results are always placed before Y25. At the top of the group, we place either the person with the earliest proven ancestor or with the strongest matches to all others in the group. All others are compared to the person at the top and are sorted numerically by the Code # in Column 1. Each distinct group is colour coded and any marker different to that at the top of the group is clear, so that the differences within the group are emphasised.
3 CURRENT
GROUPINGS
3.1 Group 1, Cork: Main Group & Associated Category
These three results here present no grouping
problems, as they are exact Y25 matches, and close Y37 matches. #8958 and
#12757 are particularly close with just one mismatch on marker 35. All three are undoubtedly
related. It transpires that all live within close proximity to each other near
Cork City in Ireland. Two of them have traced roots to farming stock in North
West Co Cork while the other’s roots are in a merchant family of Cork City.
Hitherto none of the three knew of each other or of a relationship. All are now
working on a paper trail to see if the relationship can be authenticated by
conventional means, but
despite extensive research no proven links have emerged.
#14921
of a Donegal ancestry, somewhat surprisingly has very distant links to each in Group
1. #40972 matches only #8958 and
that no match exists with #10359 is somewhat surprising, as both are known to
have ancestors in the same parish in North West Cork, demonstrating that people
of the same surname living close to each other are not necessarily related.
3.2 Group 2, Tipperary/
Kilkenny : Main Group & Associated Category
#12446 and #43219 are exact 37/37
matches. This is exciting as it links families in Virginia, USA and Victoria,
Australia with an 84% probability of having a common ancestor within the past
four generations. Neither party were aware of the other’s existence although
both understood their ancestry to be in Tipperary. This result certainly proves
the power of DNA testing in genealogy. #87583 closely matches both of the above
as does #21517 although he
has only done the Y25. All with the exception of #21517 believe their
earliest known ancestor to be from Co. Tipperary or the neighbouring Co. Kilkenny. #87583 is of a family that has been in
the Cloran (near Fethard in south Tipperary) area for generations. #21517
believes his roots to be in Dingle, Co Kerry. This grouping is interesting, as
it appears to solidly link Kerry to Tipperary and Kilkenny and future matching
results are keenly awaited.
The ‘Associated
Category’ include;
· #17203 with a proven link to the
famous Shee family of Kilkenny, one of the most prominent families there in the
period 1550 to 1700.
· #27770 with a Tipperary ancestry
· #N21890’ with a ‘Schroyer’ surname
who asked to participate in the O’Shea project and does link to the group.
· #XOUX8 from a Kilkenny background. As
his test was not done by FTDNA it is more difficult to correctly allocate him to a specific group.
3.3 Group 3, South Kerry:
Main Group, Subsets A & B and Associated Category
#89170 & #89883 are exact Y37
matches. Both were raised in New York and know each other, although up to now
were unaware of any relationship. Neither is positive of their Irish origins. Interestingly one is a ‘Shay’
with the other an ‘O’Shea’, demonstrating how surnames can change in the US.
#82020 & # 100153 (a new participant since our last
report) are also an exact match and both match the above two 36/37. #82020
thinks his ancestor emigrated from Tralee, while #100153 is unsure of his Irish
background. Again their surnames differ. These four results form a close-knit
unit, and efforts are currently underway utilising an Irish-based genealogist
to see if the links can be proven by conventional means, and if an exact Irish location
of origin can be established. #37384’s ancestors are said to be from the
barony of Tirawley in Co Mayo, an area not normally associated with O’Sheas. #40971’s ancestors are from near
Waterville in Co Kerry close to the source of the river Inny where historically
the O’Shea clan is believed to have originated. His presence in this group
suggests it may be representative of the Iveragh peninsula. #41184 is from the
Beara peninsula in south west Cork.
In subset 3A, #40970 also comes from
the Beara peninsula in South West Cork. His ancestors are believed to have
moved from Kerry in the nineteenth century, possibly to work in the local copper
mines at Allihies
In subset 3B, #N6680 is not certain
of his origins but thought it may have been Co Cork.
The ‘Associated Category’ includes:
· #34950 whose ancestor came from
Duhallow in North Cork. It was erroneously expected he would match Group 1,
again proving that participants whose ancestors came from adjacent locations do
not necessarily match genetically. He is a close relation and a close match to
#91857 and while both are well aware of their relationship, neither knew that
the other was in the project. The DNA tests accurately reflect their close
relationship with a 36/37 match; a pleasing verification that the DNA testing
process works.
· #40268 thought he might have had Cork
origins.
· #58344 is not sure of his ancestry.
The presence of people with proven
connections to the Beara and the Iveragh peninsulas, suggest that this group
may be representative of South-West Kerry. It would be particularly significant
if the group turned out to be distinct from the O’Sheas of West Kerry on the
Dingle peninsula, but participants with proven origins there, have not to-date
been forthcoming.
3.4
Group 4, North Kerry, Limerick & Clare
This is a new group since our last report, created because of matches
between participants with proven links to the three areas. The group is closely
linked to, and difficult to differentiate from, Group 3 and future results may
well substantially alter it. #15527 has a Clare background although the family is believed to have
originated in Kerry. #34730 is from Killorglin on the northern edge of the
Iveragh peninsula in Kerry. #45584’s most distant ancestor served in the
British Navy in the eighteenth and nineteenth centuries and a somewhat
discredited pedigree places an earlier generation in Co. Limerick. #83190 is
definitely from the Cahersiveen area in South Kerry, and does appear to be the odd man out
in the group. He has also close matches to group 3. #97054 is definitely from
Limerick as his parents emigrated from there in the early 1950s and he has a
documented pedigree there, back to the early 1800s.
In
the ‘Associated Category’, #80679 is not certain of his origins but believes it may be
Limerick. #84068 is unsure of his Irish origins but thought
it might be Cork or Kerry. #91300 believes his ancestor was from
Kilkenny, but test results suggest otherwise. #38756 like #15527 has a Clare
background with the understanding that the family had moved there from Kerry.
The attached Table 2 sets out a comparison of
the modal results of the four groups. This shows for instance that a new result
with a value of 14 for marker 6 is likely to be Group 1, while a value of 12
for marker 5 is likely to be Group 2.
3.5 Unassigned 1b group:
Our group selection criteria
prohibit these eight
results from inclusion in a group, subset or associated category. They could be
considered as separate groups at present containing just one member and as more
results are processed it is possible they will join new groups.
· #8927 with definite Tipperary roots
has compared his results to those of other Irish clans and indications are they
may be closer to the O’Donoghues than the O’Sheas.
· #17313 has definite Dingle, Co.
Kerry roots but his results are very different to others from Kerry indicating
some kind of anomaly.
· #29808 is unaware of his ancestry
beyond his grandfather who was raised in an orphanage in the US.
· All that is known of #33720 is that
his ancestor turned up in Nova Scotia, a popular destination for Irish from
Waterford and South Tipperary.
· #47933 one of six Seays in the
project has no close matches and neither have the other R1b Seays, #69973 &
#93677.
· #78217 believes his ancestor was
born in Tipperary but raised in Kilkenny.
3.6 Y12 Grouping
#99075 & #99085 have done only the Y12 and are noted as such pending
possible upgrades. No attempt has been made to position either in a group based
on the available results
3.7 Haplogroup I1a grouping
People with different haplogroups
cannot be related within thousands of years. Three participants #s 13686, 38052
& 46110 are found to be of the I1a haplogroup. All three have a ‘Seay’
surname and none have proven links to an Irish ancestor.
3.8 Haplogroup ‘G’ grouping
One, #10703 falls into this
haplogroup. He is not aware of his Irish origins and understands there may have
been an adoption or name change in his background.
4
THE Y-DNA PROJECT FUND
The O’Shea yDNA Surname Project
started in 2003 and has now attracted over fifty participants. This is exciting
and we have seen that some very interesting matches have occurred. However it
can be tantalising if a definite link can’t be established to a specific Irish location.
This is as frustrating as ‘hitting the wall’ when the paper trail runs out in
conventional genealogy. The project administrators accept that there are not
enough O’Sheas still residing in Ireland, involved in the project. We have
tried to attract volunteers from traditional O’Shea areas, but this has proved
difficult, as understandably, people who have lived in such areas for
generations do not feel the need to spend money to find out where they came
from! In the past, we have voluntarily funded several people to try to address
this problem, and now we ask others to help to progress our project in a way
that will eventually assist all participants.
In August 2007 we set up a fund within FamilyTreeDNA, out of which will be paid the test costs for people who
otherwise would be unlikely to join the scheme. The goal is to identify
strategic participants in areas long associated with the O’Sheas. Selection
criteria would include whether or not the participant can document a good
measure of male-line ancestry, but the administrators will also consider
funding any prospective member considered of interest by any existing member of
our project.
One member has very generously started off the fund with a contribution of $1000 and another subsequently donated $200. We are extremely grateful to them. Perhaps others might consider contributing. If so, while any donation no matter how small will be gratefully received, $200 would be very appropriate, as it would allow for the individual Y37 test that is now considered the minimum. If you would like to contribute to the fund please use: http://www.familytreedna.com/contribution.html or email one of the administrators who will help with the procedures. Any contribution will be gratefully acknowledged and details of how the fund has been utilised will be regularly circulated.
Since August the Irish-based co-administrators have actively canvassed suitable candidates. While kits have been dispensed, unfortunately none have as yet been returned. Replacement suitable candidates are being actively pursued. It is hard to understand why Kerry folk in particular are so reluctant to join the project, even when they are being funded. We will just have to be patient and to ‘make haste slowly’.
5 PHYLOGENETIC
NETWORK DIAGRAM
Such a diagram is one way of graphically
representing possible relationships based on DNA results. The simplest diagram
is a tree where there is only one path between any two nodes. The objective is
to have the tree resemble as closely as possible the most likely relationships
between the project participants. The nodes represent key individuals, either
ancestors or descendants. The lines between nodes represent the relationships.
There are usually many ways of
drawing the same tree or network. It is difficult to determine absolute
relationships, e.g. which node represents the most distant common ancestor, but
the tree or network generally gives a good indication of which nodes are likely
to be more closely related to other nodes. The shorter the path between two
nodes, the closer the relationship is assumed to be. Our diagram (produced by
courtesy of Fluxus-Engineering software) is shown
in Appendix A. It displays the current relevant Y37 and it does
tie up with the groups shown in Table 1, with the four groups showing up on the diagram
as ‘clusters’ remote from each other, while ‘unassigned results sit remotely.
6 MATCHES
WITH OTHERS
As part of the FTDNA service, participants who have
a)
Signed
the release form when originally applying
b)
Ticked
the ‘I want my matches to be set against the entire database’ under the section ‘Set-up
preferences’
are provided with a summary of matches with all
others in their database, including those with non O’Shea/Shea surnames. Appendix B, sets
out these matches, based on data downloaded on 28 Nov 2007. FTDNA do indicate that even exact matches for Y12
tests between persons of different surnames have no real significance. Close
matches on Y25 or Y37 might be significant but the table clearly shows that
such are rare.
As can be seen matches vary
considerably. In all cases none of the exact matches with Non-O’Sheas on the
Y12 test survive as an exact or even a (-1) mismatch on the Y25 proving the
initial connection was not significant. There are few Y37 matches and in almost
all cases the links are to fellow project members. The surnames with which the
O’Sheas match vary, but all look ‘European,’ with a sprinkling of Irish names
such as O’Connor, O’Donoghue, McCarthy, Sullivan etc. The figures vary
continually as FTDNA process both more O’Shea and ‘Non-O’Shea’ results. We stress that these
results, while of general interest, have no real significance.
7 SUMMARY
Almost
all recent results show linkages to Kerry, which is to be expected as it has
always been recognised as the spiritual home of the O’Sheas. We now have
sufficient data to allow us a reasonable chance of suggesting the Irish origin
of a project member with no knowledge of their Irish roots. More results,
particularly of those with proven links to an Irish location, will improve the accuracy
of our predictions.
Approx locations of the
origins of our current four groups
There is a surprising diversity in the surnames
in our project with nine variants at present as below. Of the 49 current participants, 31 are based in the US, 8
in Ireland, 4 in Australia, 2 in Canada and one each in South Africa, Hong
Kong, England and Italy. The spread of people in the US is shown on the right.
We have had an initial look
at comparing our O’Shea results with those of other Irish Clans. While viewing
findings with caution, the O’Sheas do appear to match reasonably closely, the
O’Donoghues of the Glens and the O’Sullivans. The former an offshoot of the
main O’Donoghue clan were based near Glenflesk in the east of Kerry, while the
O’Sullivans were originally from Tipperary but moved to South Kerry in the
eleventh & twelfth centuries where they displaced the O’Sheas. It would be
of particular interest to compare our results with the Falvey and O’Connell
Clans, as all three Clans are understood to be from the same Corcú Duibhne root
in West Kerry in medieval times. However neither the Falveys nor the O’Connells
have instigated surname projects. This is understandable for the Falveys, whose
numbers have shrunk over the generations, and now within Ireland have only a token
presence, mainly in Kerry and Cork. However the O’Connells are a major clan,
one of the most numerous within Ireland and thus with a large overseas
contingent, and it is surprising that a project hasn’t been instigated. Perhaps
this may change in the near future.
FTDNA now offer a Y67 test. We feel at the present
time there is little point in doing this, unless a very close match has already
been found on the Y37. For existing participants considering upgrading, we
recommend upgrading to the Y37. For new members we again recommend the
Y37. The Y12 results are too
general to be of any great use and most people undertaking this upgrade shortly
afterwards, but it is cheaper to go directly to the Y37 rather than upgrading
from the Y12.
Academic work on the genetics of
Irish clans is proceeding at a number of institutions particularly at Trinity
College in Dublin where some related articles may be accessed at www.gen.tcd.ie/molpopgen/resources.php
O’Shea Clan
Society December 2007
Appendix A
Phylogenetic
Network Diagram
Appendix B
Matches both with O’Sheas
& Non-O’Sheas as at 28th Nov. 2007.
|
Code |
Y12 |
Y25 |
Y37 |
|||||||||
|
# |
0 or -1 |
0 |
-1 |
-2 |
0 |
-1 |
-2 |
-3 |
-4 |
|
||
|
8927 |
5 |
0 |
5 |
35 |
0 |
0 |
0 |
1 |
1 |
|
||
|
10359 |
208 |
2 |
5 |
39 |
0 |
0 |
0 |
1 |
1 |
|
||
|
12446 |
29 |
2 |
8 |
10 |
1 |
0 |
1 |
1 |
1 |
|
||
|
12757 |
208 |
2 |
5 |
39 |
0 |
1 |
0 |
2 |
0 |
|
||
|
13686 |
3 |
3 |
0 |
1 |
|
|
|
|
|
|
||
|
17203 |
5 |
0 |
0 |
3 |
0 |
0 |
0 |
0 |
1 |
|
||
|
27770 |
4 |
0 |
2 |
0 |
0 |
0 |
0 |
3 |
3 |
|
||
|
29808 |
339 |
1 |
5 |
19 |
|
|
|
|
|
|
||
|
33720 |
14 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
||
|
34730 |
57 |
0 |
1 |
9 |
0 |
0 |
0 |
0 |
5 |
|
||
|
34950 |
2 |
0 |
1 |
0 |
0 |
1 |
0 |
0 |
0 |
|
||
|
38052 |
9 |
1 |
6 |
2 |
|
|
|
|
|
|
||
|
40268 |
0 |
0 |
0 |
0 |
|
|
|
|
|
|
||
|
40970 |
230 |
0 |
4 |
3 |
0 |
0 |
0 |
2 |
2 |
|
||
|
40971 |
53 |
0 |
0 |
6 |
0 |
0 |
3 |
3 |
0 |
|
||
|
40972 |
38 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
||
|
41184 |
230 |
0 |
0 |
7 |
0 |
0 |
3 |
3 |
0 |
|
||
|
46110 |
43 |
5 |
4 |
2 |
|
|
|
|
|
|
||
|
47993 |
659 |
0 |
10 |
32 |
|
|
|
|
|
|
||
|
N6680 |
165 |
0 |
0 |
14 |
0 |
0 |
0 |
0 |
2 |
|
||
|
N21890 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
1 |
|
||
|
69973 |
12 |
0 |
5 |
1 |
0 |
2 |
1 |
2 |
1 |
|
||
|
78217 |
34 |
0 |
2 |
18 |
0 |
0 |
0 |
0 |
1 |
|
||
|
82020 |
230 |
3 |
4 |
8 |
1 |
2 |
0 |
3 |
1 |
|
||
|
87583 |
4 |
0 |
1 |
8 |
0 |
0 |
2 |
0 |
2 |
|
||
|
91300 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
||
|
91857 |
30 |
0 |
1 |
4 |
0 |
1 |
0 |
0 |
1 |
|
||
|
93667 |
3 |
0 |
2 |
11 |
|
|
|
|
|
|
||
|
97054 |
165 |
0 |
3 |
18 |
0 |
0 |
1 |
1 |
0 |
|
||
|
100153 |
230 |
3 |
4 |
8 |
1 |
2 |
0 |
3 |
1 |
|
||