Help me understand DNA testing

[ForresterModern]

This is similar to my situation. I have two uncles (my mum's brothers) who are the last male O'Callaghans of the name in our family. One never married, and the other married someone who already had a daughter and didn't want more children. They are both well into their eighties, so there's no likelihood that either of them will have a genetic son now.

I would love to know what a DNA test on either one of them would show, but I am hesitant to ask them to get tested. I doubt if they would immediately realise that it would only take a scraping from under the tongue (or so I've heard). OTOH, if it is left too long, then eventually it will be too late, for the saddest of reasons. I think I ought to speak to my mum about this. She is the family genealogist anyway, so she would be interested in the results I'm sure.

The burning question that it ought to answer is whether we are all descendants of Ceallachan of Cashel, King of Munster, d.953, or simply, as we already know, descended from a long line of Irish sailors. Whether a Y DNA test could even answer that I'm not sure, but the only way to really find out is to test one of my uncles.

Theis no scientific way to prove it, even if you have yDNA profile on Ceallachan of Cashel or one of his direct patrilineal verified relatives, your paper trail is far more likely to yield results. YDNA on your uncles would be helpful for you to trace that particular lineage and see if it matches up with those of other callahans, particulalry in sizable numbers in the local where that ancestor hailed from. If one of them could prove descendency then you might be in luck, but still all you have proved is that you and the other person with the matching profile have at least one common ancestor, not necessarily that it actually is Ceallachan. It could be anyone who he shared a common male ancestry with, potentially generations above even Ceallachan---his father, uncle, nephews, grandfather, great grandfather, their other siblings or uncles or nephews, etc... All yDNA profiling proves is common ancestry on the male lineage side of SOME SORT, not any specific sort.

[wvpiper]

So how does it work if my Scottish line comes thrugh my mother, her father and grandfather, thence through her great grandmother to her GGM's father and patrilineal from there?

Obviously the Y test isn't right, but will the mtDNA test shed light any better? Or is this a case of "you're hosed, turpin"?

Male descendants of your mother's family can be tested to affirm that line. An uncle or male cousin, for example.

To the OP-try hooking up with a surname DNA project, if available. You can get the kit, get the results and compare them to other tests and get support on what it all means.

Interesting story from 60 minutes a while back.

This business of genetic genealogy is fraught with limitations

http://www.cbsnews.com/stories/2007/...n3334427.shtml

[Alan H]

This is similar to my situation. I have two uncles (my mum's brothers) who are the last male O'Callaghans of the name in our family. One never married, and the other married someone who already had a daughter and didn't want more children. They are both well into their eighties, so there's no likelihood that either of them will have a genetic son now.

I would love to know what a DNA test on either one of them would show, but I am hesitant to ask them to get tested. I doubt if they would immediately realise that it would only take a scraping from under the tongue (or so I've heard). OTOH, if it is left too long, then eventually it will be too late, for the saddest of reasons. I think I ought to speak to my mum about this. She is the family genealogist anyway, so she would be interested in the results I'm sure.

The burning question that it ought to answer is whether we are all descendants of Ceallachan of Cashel, King of Munster, d.953, or simply, as we already know, descended from a long line of Irish sailors. Whether a Y DNA test could even answer that I'm not sure, but the only way to really find out is to test one of my uncles.

There's no way to know For Sure. Even if we could retrieve a tissue specimen from some guy who died in 953 AD, and it had DNA intact enough to test the various marker haplotypes, you still wouldn't KNOW.

Even if you DID have that tissue, and there's been direct partilineal descent during that 1,000 yeas, well.... it's been over a thousand years. Random mutation could have introduced anywhere from 4 -40 point mutations during that period. Genetic Drift changes haplotypes.

If you have a large number of records of people who can without a doubt PROVE direct patrilineal descent from that EXACT MAN...and then get yDNA samples from those people to establish a pool of genetic variation, well, then you have a start. You'd then apply various parsimonious models of change to the array of mutations that you've got in these proven descendents to come to some sort of "hopeful best fit" hypothesis about Ceallachan of Cashel, King of Munsters actual genotype.

You could then attempt to calculate the probability that your Uncle's haplotype array could be derived from Ceallachan of Cashel, King of Munsters theoretical array.

There's something VERY, VERY important here, which gets glossed over a lot. It deserves its own post.

[Alan H]

All, and I mean ALL of the various models that outline lineage depend on a series of mathematical models which predict/model genetic change. EVERY SINGLE ONE of these models employ various algorithms which put into practice a common principle....that of "parsimony".

Parsimony basically says that the simplest series of events that result in a measured or observed outcome is the most likely series of events that created that outcome.

What parsimony does is calculate the simplest series of events....a mutation here, followed by that mutation, etc. etc.

But just because it calculates the SIMPLEST PATH to an outcome does not in any way guarantee that is what actually happened.

[pdcorlis]

I think your paper trail is pretty freaking good, if you really have that information documented...

snip...

If you really wanted to pursue this route, I'd research mtDNA, instead and take a look at Moms side of the family.

snip...

Thanks Alan - yeah, there are a few advantages of having a super rare last name (Corliss). I did most of the work on dad's side but I have had a couple of hard core genealogists look it over and they concur. Mom's side (Paskett)is also solid back to the early 1600s thanks to a few generations of hard working Paskett genealogists. I can't imagine doing this before computers, those folks must have had the patience of a saint.

I will certainly look into mtDNA - thanks for the advice.

Phil.

[Alan H]

Honestly, the whole topic is phenomenally complex. It works pretty well over huge amounts of time. The first bits of research, conducted not on sequences but on DNA hybridization experiments, done on nuclear DNA, suggested that all of us are descended from a common pool of women (not ONE woman, a common, small pool, possibly an actual working community) based in central East Africa, around 250,000 years ago.

As some point when this was publicised, another group attempted a similar study based on yDNA and got similar results. The vitally important fact here is that what the geneticists came up with agreed well with what the classical anthropologists had figured out.

Over spans of time like that, things like mutation rates tend to even out. Mutations, the events which result in haplotypes, do not occur like the ticking of a clock. They are random, though how the natural environment selects them is very much NOT random, at all, at all. There may be very long periods of time before a mutation occurs which is not lethal, or which confers no special evolutionary advantage or disadvantage. Then again, you might get one today, and then you might get another one, tomorrow. It's RANDOM.

Try to "set a mutation clock" by estimating mutation rates over the span of a few hundred years is shaving the time interval in which these random events occur, to too small of an interval to actually accurately calculate a rate.


Do not ask DNA testing to determine if you are related to King MacSchnozzle from the 14th Century.

You CAN look into DNA testing to get a rough estimate of how closely related you are to other people with the same surname, though remember that the correlation between the continuance of surname and the passing on of haplotypes is not 100%. You don't know if your ancestor in 1720 was ADOPTED, for example. There are hundreds of example of people changing names throughout history.

One of my clans, the Hall Border Family group/Clan is in fact supposed to be descended from a Norman family, the FizWilliams. A Younger son took the name "Hall" to distinguish himself from his older brother, who kept "Fitzwilliam". So if for some reason I wanted to document ancestry earlier than 1009, I'd have to go looking for haplotypes in people named "FitzWilliam".

But what if I didn't KNOW, that the younger son changed his name? HMMMM????? I'd be barking up the wrong tree.

DNA testing is NOT....NOT....the Be-All, End-All of geneology, and outside of perhaps 2-3 generations back, it flat-out ***PROVES** nothing about ancestry. It gives clues, hints and suggestions, but it does not replace a paper trail.

[Alan H]

Thanks Alan - yeah, there are a few advantages of having a super rare last name (Corliss). I did most of the work on dad's side but I have had a couple of hard core genealogists look it over and they concur. Mom's side (Paskett)is also solid back to the early 1600s thanks to a few generations of hard working Paskett genealogists. I can't imagine doing this before computers, those folks must have had the patience of a saint.

I will certainly look into mtDNA - thanks for the advice.

Phil.

Honestly, Phil, you've got a paper trail on your mothers side that goes back to the 1600's as well? No kidding?

Phil, you're done. DONE. See my previous post. mtDNA testing will give you suggestions about general areas, and general haplotypes that mom might be connected to. It will NOT tell you that your ancestor in 1537 lived in Paris, on 445 Rue De Genome Street.

[pdcorlis]

Phil, you're done. DONE...

I suspected that Alan, but I see so much about DNA in genealogy I wondered just what I may be missing.

As an aside, I have been lucky enough to stand on the same streets as my ancestors, sit in the same churches where they were married and buried some 500 years ago. It is a surprisingly moving and powerful experience.

I hope some day everyone who does genealogical work can share it.

[O'Callaghan]

There's no way to know For Sure. Even if we could retrieve a tissue specimen from some guy who died in 953 AD, and it had DNA intact enough to test the various marker haplotypes, you still wouldn't KNOW.

Even if you DID have that tissue, and there's been direct partilineal descent during that 1,000 yeas, well.... it's been over a thousand years. Random mutation could have introduced anywhere from 4 -40 point mutations during that period. Genetic Drift changes haplotypes.

If you have a large number of records of people who can without a doubt PROVE direct patrilineal descent from that EXACT MAN...and then get yDNA samples from those people to establish a pool of genetic variation, well, then you have a start. You'd then apply various parsimonious models of change to the array of mutations that you've got in these proven descendents to come to some sort of "hopeful best fit" hypothesis about Ceallachan of Cashel, King of Munsters actual genotype.

You could then attempt to calculate the probability that your Uncle's haplotype array could be derived from Ceallachan of Cashel, King of Munsters theoretical array.

There's something VERY, VERY important here, which gets glossed over a lot. It deserves its own post.

There are undoubtedly people who claim descent from this individual and have a family tree showing that. They of course include the clan chief, who lives in Barcelona, Spain, and they include Baron Lismore of Lismore, Co. Waterford. Also the McCarthy Mor (clan chief of the Macarthys). I don't know if the DNA of any of these people or their close relatives is recorded, and of course we can't get DNA from him,as you point out.

Traditionally, according to the old lineages, everyone whose name is Ceallachain, O' Ceallachain, Callahan, Callaghan or O' Callaghan is descended from him, not to mention all the Sullivans and O'Sullivans, and the MacCarthys and McCarthys as well! The Macarthys are actually the senior line of descent from Ceallachain of Cashel, and the Callaghan line was founded by a grandson who also bore the name Ceallachain.

Some of the Macgraths, McGraths and Kennedys would also be descendants of Ceallachain of Cashel, but those names also have other known separate origins, as Irish surnames are based on Gaelic names that were used as single names before surnames existed, and there were multiple individuals called Maccraith and Cenneidhe in entirely separate tribes who left descendants. No, I don't know which one the Kennedys of Hyannisport (and Co. Wexford) descend from, but I think they know, I understand that they have a very extensive family tree plotted out going a very long way back.

However, it is also said that some clan members were not blood relatives, but their descendants also took the name of the clan, which is the same as the name of its founder, as their surname. So that's where you get people of the same name who are not related. Even then, some (most?) of them would have intermarried with members of the clan who were already related to its progenitor, so that would tend to make many of their descendants related to him anyway.

Putting this in perspective, I read somewhere that there are 100,000 Callaghans in Co. Cork alone. This does not necessarily even rule out the possibility that after 1,000 years perhaps all of them might be descendants of Ceallachain of Cashel. If everyone in Europe is a descendant of Charlemagne, as some maintain, then it seems possible, although I haven't attempted to do any calculations to see if it's feasible.

[mudd]

i got mine done a few months ago and have just the past week got all my final results in from my ydna and mtdna test and ...

I find this very interesting information. (From someone who can find only a limited hereditary paper trail.)

If you don't mind saying, where did you have this done?