Thursday, 31 March 2016

The Birth of a New Genetic Family (R1b-GF6)

If you have read previous posts in this blog, you will have seen that I frequently predict the emergence of new genetic families. Farrell (and its variants) is likely to be a multi-origin surname that sprung up at various times in disparate locations throughout Ireland (and maybe even in Scotland and England). And as more people test and join the project, some of the project members who are currently in the Ungrouped section will suddenly find that they have a new match, a genetic cousin, and they are no longer "all alone in the world" but rather have connected with genetic family.

There follows a case in point, where two individuals tested recently, to see if they were indeed related as their paper trails suggested. And the results came back positive. They are related. And not only that, but they are also related to other members in the Ungrouped section. All four of them have started a whole new genetic family, the first for 2016. And we will call this new family: R1b-GF6.

Here is their story.

New Member -4174, Farrell, MDKA Longford

Basic information

Kit Number ending: -4174

Name: Farrell

Current Group: R1b-GF6

MDKA: Daniel Joseph Farrell, b abt 1825 Longford, d 1901 Iowa

Background: His family folklore and Census data research says that his great-grandfather was born about 1825 in Longford, Ireland, and came to the US in about 1857. There are Census records that suggest this, but there is some uncertainty about their being the correct ones. This new member was contacted by another Farrell (see -4178 below) who thought that they might have a common great grandfather. Both men did the Y-DNA-37 test to resolve this question. Here is the Farrell pedigree of the present individual:

1) Daniel Joseph Farrell b.15 Aug 1825, Longford, County Leinster, Ireland (or Cork); d.7 Jul 1901, Dunlap, Iowa
2) Daniel Joseph Farrell b.29 Jan 1860, Illinois; d.31 Oct 1934, Omaha, NE
3) Andrew Harry Farrell, b.18 Mar 1910, Omaha, NE; d.15 Jan 1993, Fairfax, VA, USA
4) Researcher: PDF (pfarrell at pfarrell dot com)

Other projects: in accordance with our recommendations (see Getting the Most out of your DNA Test), this member should join several appropriate haplogroup & geographic projects including the following:

STR marker profile

This individual did the Y-DNA-37 test and at the 37 marker level he has 37 matches with a lot of Irish sounding names like Hart, Hogan and O’Brien ... and one Farrell - the presumed second cousin with known ancestry in Ireland (see below). This makes it highly probable that his particular direct male line is Irish as suspected.

His “Genetic Distance” (GD) to his presumed cousin is 3/37. This is a relatively close match rather than a very close match, and raises a question about the closeness of their relationship. Are they really second cousins or is the relationship further back than that (third, fourth or fifth cousins)? Nevertheless, the Y-DNA results confirm that they are closely related to each other and this match sees the birth of an entirely new genetic family within the project.

Next Steps: One way to answer the question regarding the closeness of their relatedness would be to do autosomal DNA testing (FTDNA's Family Finder test, $99). Second cousins would share about 3.125% of their atDNA in common, 3rd cousins would share about 0.78% in common, and 4th cousins would share about 0.2% in common. The Family Finder test would be able to give a better probability of the closeness of their relationship than the Y-DNA test.

Interestingly, this new member's next closest matches within the project are two men called Frawley (kits 121852 and B1778). He has a GD of 5/37 to both of these men and this is just below the threshold for declaring a match. This could be a chance finding, or it could indicate that these two additional individuals could also be part of this new genetic family. This is explored further below.

SNP marker profile

The other big clue to this persons ancestral origins is the terminal SNPs of his matches. His 37-marker matches have the following terminal SNP markers:
FGC5659 (x1), FGC5628 (x7) and L226 (x7)
A similar profile emerges when reviewing his 25-marker matches (with the addition of DC1 and YFS231286). This gives us the following estimated SNP Progression:

R- ... P312 > L21 > DF13 > ZZ10-1 > Z253 > Z2534 > L226 > FGC5660 > FGC5628 > FGC5659 > DC33 > DC1 > YFS231286

L226 is the defining SNP marker for the Irish Type III haplotype. This is a genetic signature associated with Irish surnames from the southern part of Ireland and most frequently in the counties Clare, Tipperary, & Limerick. There is a wealth of information provided by Dennis Wright, the Administrator for the Irish Type III Project, on the project's website.

Marker L226 is also the supposed genetic signature of the famous Irish High King, Brian Boru. In fact, on the Big Tree, the YFS231286 SNP marker leads us down to three individuals called O'Brien (far left below). So it looks like this particular Farrell line goes back to the ancient tribe of the Dal Cais. They may even have fought at the battle of Clontarf in 1014! Who knows, maybe that is where they got their Farrell name (Farrell means Man of Valour in Gaelic). As more people join this Farrell group and do SNP testing, we should be able to clarify this interesting association.

Terminal SNP markers of this individual's matches (click to enlarge)

Next Steps: We could confirm that the SNP Progression is correct by doing sequential single SNP tests (at $39 each) starting with (say) L226 and working downstream toward YFS231286. Or we could test with one of the available SNP Packs - the best bet probably being the Z253 SNP Pack ($119) which includes all the downstream SNPs in our estimated SNP Progression.

But do the two possible Frawley matches (kits 121852 and B1778) have a similar SNP Progression? This is discussed below.

New Member -4178, Farrell, MDKA Ireland

Basic information

Kit Number ending: -4178

Name: Farrell

Current Group: R1b-GF6

MDKA: Peter (Patrick?) Farrell born c1830, died 1902 in Kilmallock, Co. Limerick.


1) Peter (Patrick?) Farrell. Born abt 1830. Peter (Patrick?) died in Kilmallock (Limerick) in 1902; he was 72.
2) John Farrell. Born in Aug 1871 in Shandrum, County Cork, Ireland. John died in Maiden Hall, Ballincolly, County Cork, Ireland, on 21 Nov 1956; he was 85. Buried in Cooline Cemetary.
3) John (Jack) Farrell. Born on 15 Aug 1912 in Maiden Hall, Ballincolly, County Cork, Ireland. John (Jack) died in Mallow Hospital. Buried Charleville in Holy Cross Cemetary, Charleville, County Cork, Ireland.
4) JPF (DNA kit -4178). Born in Ardnageehy, Charleville Co. Cork Ireland.

Background: The following (privatised & paraphrased) account is from -4174 above.

"We think that -4178's GGrandfather Peter (Patrick) Farrell born about 1830 is the brother of my
GGrandfather Daniel J Farrell born about 1825. Our folklore family history includes a letter dated about 1900 from Peter to Daniel. And the folklore have three siblings, Daniel, Andrew and Peter. We have no written records of the parents of Daniel, Andrew and Peter. There are US Census records for Daniel in the 1880 and 1870 records."

Other projects: in accordance with our recommendations (see Getting the Most out of your DNA Test), this member should join several appropriate haplogroup & geographic projects including the following:

STR marker profile

Like his presumed Farrell cousin above, this new project member has also taken the Y-DNA-37 test. He too has 37 matches at the 37-marker level, his one Farrell match being the individual discussed above.

The Genetic Distance (GD) between him and his cousin -4174 is 3/37 (as already described above) but the GD to his next two nearest matches (the Frawley matches; kits 121852 and B1778) is 7/37 in both cases, indicating that he is possibly more distantly related to the two Frawley matches than his presumed Farrell cousin.

It is helpful to create a matrix of these 4 individuals and the GD between each of them in turn:

The Genetic Distance between all 4 individuals

Although the genetic distance suggests a tantalisingly close relationship, it is not conclusive and we need to look for other evidence of a close connection between these four individuals.

Incidentally, the two Frawley matches have both tested out to 67 markers and are a GD of 9/67 from each other (again, just outside the threshold for declaring them to be a match to each other).

SNP marker profile

Not surprisingly, both presumed Farrell cousins (-4174 & -4178) have a similar STR and SNP marker profile. This individual's matches have almost the same profile of terminal SNPs as his presumed Farrell cousin:

  • at 37 markers: L226 x8, FGC5628 x5, FGC5659 x1, YFS231286 x1
  • at 25 markers: L226 x4, FGC5628 x2, DC1 x1 x1, YFS231286 x3

Thus, this is consistent with our previously estimated SNP Progression:
R- ... P312 > L21 > DF13 > ZZ10-1 > Z253 > Z2534 > L226 > FGC5660 > FGC5628 > FGC5659 > DC33 > DC1 > YFS231286

Both Frawley matches have tested out to 67 markers. Surprisingly, the first Frawley match has only 1 match himself and that is at 12 markers only, and it is to another Farrell (N70429, who in turn has only tested to the 12-marker level and has not tested any downstream SNPs). The second Frawley match (B1178) has the following profile:
  • At 67 markers, 17 matches with terminal SNPs: DC30 x1, FGC5628 x3, L226 x7, YFS231286 x1
  • At 37 markers, 4 matches: DC30 x1, L226 x1
  • At 25 markers, 31 matches: DC36 x1, FGC5628 x1, L226 x5

Thus the second Frawley profile is largely consistent with what we had already predicted to be the SNP Progression for our two new members, the presumed Farrell cousins -4174 & -4178. But the clincher is the fact that this second Frawley match (B1178) has already done sequential single SNP testing and is positive (in green below) for the following SNPs:
DF13, L226, FGC5628
However, he has also tested negative (in red below) for the following SNPs:
FGC5659, DC1
Testing for DC40 is ongoing (in orange below; this SNP is not currently placed on Alex Williamson's Big Tree).

The second Frawley match (-1178) is not positive for SNPs below FGC5628 (green)

So here we have a contradiction. The second Frawley match does NOT completely fit with our estimated SNP Progression for the two new members. He matches as far down as FGC5628 but not further downstream than that. This means either that:
  1. our predicted SNP Progression is correct as far as FGC5628 but incorrect below this
  2. or that the SNPs further downstream than FGC5628 are correct for the Farrell individuals but not for the Frawley matches. In other words, maybe they branched away from each other and the SNPs downstream of FGC5628 represent more recent developments on the Farrell branch of this particular portion of the human evolutionary tree (e.g. since the divergence of the surname Farrell from Frawley); 
  3. or that the two presumed Farrell cousins and the two Frawley matches are not related at all and what we are seeing is an example of Convergence - in other words, the genetic match has occurred by chance alone (more or less) and they just happen to have similar surnames. 

Consulting the YFULL Tree reveals that the age of FGC5628 is approximately 1350 years old (95% Confidence Interval 1600-1150 years before present), and DC1 formed approximately 1100 ybp (95% CI 1350-850 ybp), so this does not rule out option 2 above as a possibility.

But let's look for one more piece of evidence ...

Rare Marker Values

By consulting Leo Little's excellent chart on the distribution frequencies of the different values for each STR marker, we can identify if there are any potentially rare marker values. If there are, and they are shared among the 4 individuals, then this lends support to the possibility that they are all relatively closely related to each other and probably shared a common ancestor since the emergence of surnames about 1000 years ago.

And the good news is that there do indeed appear to be shared rare marker values among these 4 individuals, justifying their being grouped together in the same genetic family, namely R1b-GF6.

Here are the shared rare marker values among the 2 Farrell cousins and the 2 Frawley matches:
  • the values for DYS459a & b (the 14th & 15th STR markers) are 8 and 9 (written 8-9). These values occur in approximately 3% and 17% of the general R1b population and thus the first of these markers can be considered rare. In fact, these 4 individuals are the only R1b individuals in the entire Farrell project to have this sequence of values, and no other R1b project member has a value of 8 for DYS 459a.
  • the values for the DYS464a-d markers are 13-13-15-17. These values occur in approximately 2%, 1%, 11%, and 69% of the general R1b population and thus the first two can be considered rare. Nobody else in the entire project has a value of 13 for DYS464b and only 1 other person in the entire project has a value of 13 for DYS464a.

The only caveat is that these particular rare marker values are quite common below L226, in fact they are the norm (see the DNA Results page for the L226 Haplogroup Project). So although we can be reasonably confident that the members of R1b-GF6 lie somewhere below Z253 and L226, where they lie below that is still open to question. And the further downstream we go from L226, the more speculative our "best guess" becomes.

Nevertheless, on the basis of these rare marker values, and taking into consideration the other evidence, it seems reasonable to conclude that these 4 individuals are indeed related (within a genealogical timeframe) and should be grouped together.

Shared rare marker values (red arrows)

The biggest advantage of grouping these four members together is that it provides further evidence of the ancestral origins of these people. One of them has a Most Distant Known Ancestor (MDKA) from Longford and another from Limerick. So on the basis of this information alone, we can say with a fair degree of confidence that the ancestral origin of group R1b-GF6 is Ireland.

Conclusions & Next Steps

So in this particular instance, the testing of the two Farrell cousins (-4174 & -4178) confirms that they are indeed related on their Farrell line. 

A GD of 3/37 is a bit surprising for second cousins but can happen nevertheless. To confirm if they are indeed second cousins, they could do an autosomal DNA test. We would expect them to share approximately 3.125% of their DNA in common if they are 2nd cousins.

As the two Farrell cousins are closely related, either of them could do confirmatory SNP testing via the Z253 SNP Pack (i.e. no need for both of them to do it). This would confirm whether or not the Farrell cousins are positive for the more downstream SNPs that at least one of the Farley matches  (-1178) tests negative for. It would also be useful if the other Farley individual -1852 did the Z253 SNP Pack - the results might suggest that the negative tests were specific to the particular Frawley individual tested (i.e. -1178) rather than all people called Frawley.

All 4 members should join the following haplogroup & geographic projects:

The Genetic Distance between the Farrell cousins and the Frawley matches was just outside the matching threshold but the estimated SNP Progressions were suitably similar to suggest that there might be a closer genetic connection than one might initially suspect. The presence of shared rare marker values lends considerable support to this theory and warrants the 4 individuals being grouped together. Further SNP testing (as indicated above) is likely to help justify this grouping.

This is a great example of how members can be waiting in the Ungrouped section for months or years  before a close match comes along. It also illustrates that once that new match arrives, a new genetic family is created. Over time, the proportion of people in the Ungrouped section should diminish and the proportion in Groups should increase.

This case also illustrates how essential it is to have precise genealogical information about the MDKA (Most Distant Known Ancestor), in particular their earliest known location (such as their place of birth). Knowing this information anchors the particular genetic family to a specific ancestral homeland and everyone in the group benefits from this knowledge. This is particularly relevant for diaspora populations who may not have documentary evidence or family lore of origins in a specific country.

No comments:

Post a Comment