DESMOND DNA PROJECT
(http://www.familytreedna.com/public/desmond)

My Personal DNA Results

Genetic Genealogy is a fast evolving and scientifically complex field, and while I am merely an informed neophyte, I can offer a brief summary of what my personal results indicate so far.

HAPLOTYPE

A Haplotype is a person's collection of STR (Short Tandom Repeat) markers. STRs mutate, therefore they are most useful for analyzing more recent genealogical connections in the past hundreds of years.

I have no exact (or close) matches in the Family Tree DNA database at 37 markers. The reason for this is some of my haplotype values are uncommon when compared to the Atlantic Modal Haplotype (AMH).

The AMH is a collection of STR values that are most common among the R1b Haplogroup. The closer your own haplotype is to the AMH, the more close matches you will find. Conversely, the more uncommon STR values you have, the less matches you have.
In my case, 10 of my 37 marker values do not match the AMH. My uncommon allele markers are 394=15 (6% of R1b have this value), 458=16 (18%), 607=14 (16%) / 439=11 (15%), 447=24 (19%), 442=13 (10%).

Here is the comparison of my results with the AMH:

(NOTE: "Fast mutating" alleles are in blue. Differences from the AMH are in italic. Differences from the target are in red.)

Here are my results for an additional 41 markers (30 FTDNA, 8 SMGF*, 3 others+):

Overrall, in my case, 16 of my 78 marker values do not match the AMH.

HAPLOGROUP

A Haplogroup is a determined by a person's SNP (Single Nucleotide Polymorphism) markers. Once formed, SNPs very rarely mutate, therefore they are most useful for determining ancient population groupings. Everyone who has the same SNP is related tens of thousands of years ago.

Based on similarities to other STR results in Family Tree DNA's database, and the Haplogroup Predictor, it was very probable that I was in the R1b haplogroup, which is defined by the presence of the M269 SNP. To confirm this, I took an additional test that confirmed that I do carry the M269 SNP.

About 2/3 of Europeans are in the R1b haplogroup. Unfortunately, there is not yet a clear distribution pattern to R1b, so it is not possible to identify an "Irish" R1b from a "English" or "French" or any other R1b. Recently however, several new SNPs (S21 and S28) have helped to break up the large R1b group into sub-clades. While these SNPs are found only in a small minority in Great Britain, they may be useful for differentiating more recent "invaders" such as Anglo-Saxons, Danes, Normans, etc. from "native" Britons.

[UPDATE (July 2007):

I was reported by EthnoAncestry as being S21+ in early 2006, which was somewhat of a surprise, since the present-day distribution of the S21 SNP seemed like it might be centered in the region of "Frisia", or the larger region encompassing Holland, Denmark and northern Germany. This result hinted that my ancestors were not originally be from Ireland.

However, as time passed and more SNP results became available, it became clear to me that this result might be in error for a number of reasons:
* No other "native" Irish were reported as S21+.
* I was NOT DYS492=13, and that value appears in a high percentage (90%+) of S21+ people.
* A few of my closest semi-matches at 67 markers had been tested as S21-.
I asked for a re-test of my sample by EthnoAncestry, and as I suspected, came back S21- this time.
The conclusion to draw from this episode is that when a genetic result seems in doubt, ask to have it re-checked.]

DESMONDs

While I am currently the only Desmond in the FTDNA database, there is are two Desmond in the Sorenson database.

Here is the comparison of me with Desmond #1:

It is a certainty that we are NOT related, because we only match on 19/28 markers, which is far too many differences for a genetic connection. Unfortunately, this Desmond [earliest ancestor is Michael, b. 1863 in London, England] does not list an ancestral location in Ireland, and since the Sorenson database is anonymous, he cannot be contacted for further info.

Here is the comparison of me with Desmond #2:

This Desmond #2 has a very interesting haplotype, as it is nearly a perfect match with others in a cluster of southern Irish people. Many of those people carry surnames connected to a tribal group known as the Eoganacht, and the Desmond surname is mentioned briefly (although cryptically) in the Eoganacht genealogies.
[I've posted more about the Eoganacht here.]

My own results match Desmond #2 and the southern Irish genetic cluster somewhat imperfectly, leaving open a number of possibilities to explain that fact:
a) The Desmonds as a group are closely related to this cluster, and I just happen to be an outlier to the group.
b) The Desmonds as a group are closely related to this cluster, and my similarity to it is just a statistical illusion with no genetic connection.
c) The Desmonds as a group are distantly related to this cluster with markers similar to mine, and Desmond #2 is the result of a non-paternal event with someone carrying a Eoganacht surname.
d) The Desmonds as a group have no common genetic connection.

Without a larger group of Desmond results to compare, no determination can be made, although I lean towards answer "a" at this time, because at 67 markers, my closest matches are still with the southern Irish cluster.

FITZGERALDs and MACCARTHYs 

I had hoped to be able to make some comparisons between the Desmond results, and the Fitzgerald and MacCarthy results. And in fact, I have found a Fitzgerald who is a 25/28 to me, and a MacCarthy who is a 22/25 match to me. Unfortunately, they are an exact 21/21 match to each other!

It would be very revealing if those two people someday expanded the number of tested markers to see if the match held up. Its entirely possible they are either both genetically either a MacCarthy or a Fitzgerald. This would not be an uncommon situation, as both families inhabited the same area in Ireland for many centuries.
(If I had to choose, I would say they are both MacCarthys, since when their markers are compared against the Sorenson database, the Fitzgerald has an exact 28/28 match with an O'Keefe, and 26/28 matchs with a Sullivan and a Callahan, who are all reportedly part of the same ancient Irish genealogical line as the MacCarthys.)
[I've posted more about the Fitzgeralds here.]

Here is the comparison of me with the Fitzgerald (from Sorenson) [earliest ancestor is John, born 1820 in County Tipperary, Ireland] and McCarthy (FTDNA 8406 / YSearch 87B4Z) [earliest ancestor is Eugene, born 1832 in Dunmanway, Ireland]

For more results of these other surnames, see the McCARTHY project and the FITZGERALD project.

The O'DONOGHUE project and SULLIVAN project may also be useful to review, as they have many participants, and are supposedly part of the same ancestral line as the MacCarthys, branching off after 400 A.D. and 600 A.D. respectively.

GEOGRAPHY

It seems a likely bet that my ancestors are from somewhere in the British Isles, since on my 25-marker comparison against the entire Family Tree DNA database, I have ONLY these close (22/25 or better) matches (as of 1/07):

England 11, Ireland 13, Scotland 6, United Kingdom 6, Wales 2
France 1, Germany 3, Sweden 1

The one caveat about the FTDNA database is that it is HEAVILY skewed towards subjects who originated in Great Britain, so maybe these matches are misleading. For instance, there are currently 6373 English, 2934 Irish, 2546 Scottish, 2290 GB/UK, 430 Welsh samples. But there are only 646 French, 486 Italian, 2337 German, 537 Spanish, and about 500 Scandinavian and 400 Benelux samples (as of 11/05).

(The importance of a geographic match is that it can be more likely than a surname to lead to a genealogical connection.
For instance, if you know that your Schmidt family can be traced back a small village in Germany, and you match a Mueller whose family can be traced back to the same small village, I'd bet money that there is some genealogical connection.)

Early on, I noticed a possible connection to a specific area in England, based on the fact that I had 10 exact 12/12 matches in the YSeach database, 6 of which were to Davenport, and 1 to Harvey.
2 of those Davenports and the Harvey listed Warwickshire, England as their place of origin. I also noted that Berkshire, which is only 2 counties south of Warwickshire, is the claimed birthplace of Gerald of Windsor ('Geraldus de Windesora'), ancestor of the Fitzgeralds.
More recently, those Davenports posted extended results in YSearch, 3 of them with 37 markers. When compared to me using those markers, their closeness drifts away, with all three of them at least a genetic distance (GD) of 11 from me.
The lesson from this story is that 12-marker results are a poor bedrock to base any accurate conclusions.

ODDS AND ENDS

Convergence - It is crucial to keep in mind that even if you match someone else's haplotype, it DOES NOT mean you are genetically related. This is because of what is known as "convergence", where the mutation of STR values mutate to common numbers. For instance, if my distant ancestors had a haplotype of 10 11 12, and yours had a haplotype of 12 11 10, but over the last couple of thousand years, mine mutated to 11 11 11, and yours mutated to 11 11 11, it would look like a match when there is really none.

Ireland & Spain - There seems to be some connection between Iberian (i.e. Spanish and Portuguese) haplotypes and Irish haplotypes. Some people have jumped to the conclusion that this proves ancient Irish myths about the "Milesians" who left Spain to populate Ireland.
But it is more likely that the reason for this connection is due to the last "ice age", which is referred to by scientists as the "Last Glacial Maximum" (LGM). The advancing ice of the LGM pushed the inhabitants of Europe back into pockets of habitation in Spain and perhaps Italy. When the bulk of Europe was repopulated after 15,000 BC, any identifiable R1b patterns were jumbled up in the re-expansion.

MY mt-DNA MATCHES

Mitochondrial DNA (mt-DNA) is passed down the maternal line.

My earliest known female ancestor is from County Mayo, Ireland in the mid-1800's. I do have one known exact match, whose ancestor traces to County Down in the same time period. Its hard to judge the geneological closeness of that match however, because dating mitochondrial DNA is difficult.

I am in haplogroup J. This haplogroup is considered a good indicator of the spread of Neolitic farmers from the Middle East into Europe in the last 10,000 years. It's interesting that haplogroup J made it all the way to Ireland, but this can be explained by the general rule that wives move into their husbands home, and over thousands of years, this allowed a Middle Eastern lineage to spread to the remotest corners of Europe

I have 12 nucleotides which differ in the hypervariable regions (HVRs) of the Cambridge Reference Sequence (CRS).
HVR1: 16063C, 16069T, 16093C, 16126C
HVR2: 73G, 228A, 263G, 295T, 309.1C, 315.1C, 462T, 489C

Here are my closest matches in the MITOSEARCH database using those 12 nucleotides

Here are my closest matches in the SORENSON database using those 12 nucleotides

MY Y-DNA MATCHES

Besides the proprietary FTDNA database, there are a few other databases available for searching.

Here are my closest matches in the YSEARCH database using 67 markers and 37 markers and 25 markers and 12 markers

Here are my closest matches in the SORENSON database using 43 markers and 34 FTDNA markers

Here are my closest matches in the YHRD database using 10 markers and 9 markers

Here are a few oddball searches:
* the YSEARCH matches for my FTDNA 38-67 markers
* the YSEARCH hits for my 9 non-AMH 1-37 markers [plus 388, 426] (not necessarily close matches)
* the YSEARCH hits for my 5 non-AMH 38-67 markers [plus 388, 426, 450] (not necessarily close matches)

First posted: 11/11/05
Last revised: 7/8/07