The Y-chromosome was used to compare living descendants of Sally Hemings with her “owner” and owner, Thomas Jefferson.
The counter-argument is that Jefferson’s other relatives in the immediate area - father, uncle (father’s brother), or cousins could be the source of the common Y-chromosome. However, there’s no evidence they were closer to her than Tom.
I suspect that, for looking as far back as Richard III or Thomas Jefferson, nothing but the Y chromosome (or, for a female ancestor, the mitochondria) would be of any meaningful use. Everything else would probably be about an equally good match for anyone in the general geographic region. It’s not a matter of how quickly the chromosomes mutate: The mutation rate for any chromosome would be slow enough to not be a problem over mere centuries, and I’m not sure that the rate for the Y is even actually any different from any other one. The real issues are assortment and crossing over.
For every one of the 23 chromosomes, you have two, one each from your mother and your father. The one from your father might have been the one from his mother or the one from his father, chosen randomly, so if you compare, say, your chromosomes #3 with your grandfather’s, there might or might not be a match. For someone as close as a grandfather, it’s overwhelmingly likely that there are going to be at least some chromosomes that do match, but as you go further back, it becomes less and less likely. And in most cases, there will be no record of which one you’d expect to have gotten, so you can’t even tell if a lack of match means anything.
The one exception is the Y chromosome, because that determines something that’s very obvious and known to the record books. If you’re a man, you have a Y, and you know that it came from your father, not your mother, and you know that he got it from his father, not from his mother, and so on.
And actually, it’s even worse than that for most chromosomes, because you’re not necessarily dealing with a complete chromosome. The two chromosomes in a pair can cross over, swapping the corresponding genes, resulting in a new pair of chromosomes that doesn’t actually match any single chromosome in any single ancestor. Again, this happens with every chromosome pair except for an X-Y pair, because they mostly don’t have corresponding genes.
That is not how it works.
If you compared your chromosomes in detail to your grandparent it would not be an either/or proposition. A portion of one of each pair of your chromosomes would match a portion of your grandparent’s chromosomes.
The chromosome #8 you get from your father is an amalgam of the chromosome #8 he got from his mother and father. In the process of meiosis when your father was producing the sperm that carried the genetic information that became a part of you, there was a crossing over, creating a chromosome #8 that is not identical to either of your grandparents. Indeed each somatic chromosome normally has many such cross overs during gametogenesis. The same applies to all the somatic chromosomes.
There is even a very limited amount of crossing over between the X and Y chromosome as a small portion of the Y is homologous to the X.
It can be harder to trace back somatic cell matches over many generations to any degree of certainty. Thus Y chromosome comparison may be favored for examining the paternal line of decent and mitochondrial comparison for the maternal line of descent.
But it would not be particularly difficult to compare Danney Williams’ DNA to a sample from Chelsea Clinton or Roger Clinton to assess a purported familial genetic relationship. A 1/4 match with Chelsea would be consistent with a half sibling relationship as would a 1/4 match with Roger be consistent with a half-uncle to nephew relationship.
Test enough allele sites and the chance of a false positive would be remote enough to establish a genetic relationship to the satisfaction of any reasonable legal standard. A negative result would be clear and convincing. The odds of a false negative are too remote as to be considered.
No, for Richard III they used mitochondrial DNA, which is passed down more or less unaltered from a mother to all her children. So they matched the skeleton’s mtDNA to the mtDNA of Richard’s sister’s daughter’s daughter’s daughter’s (etc) son.
Me too. They’d have to use the Ys, as if the skeletons are the Princes in the Tower, their mtDNA wouldn’t match Richard’s. The problem there, though, is that there’s a certain amount of speculation that the boys’ father, Edward IV, might have been illegitimate - in which case his Y chromosome (and the boys’) wouldn’t match Richard’s either way.
I don’t think they know for sure whether those skeletons (the younger one, anyway) were male or female, though.
But what about those freaky “chimera” people? Where the DNA for their fingernails IS different from some other body part?
As I understand it…those folks are rare, but suprisingly not rare rare. Problably something like triplet rare.
Yeah, as far as I know, no one knows the actual prevalence of chimerism. It makes me wonder about paternity tests. I mean, there have been cases where a woman was proven not to be the genetic mother of a child that she conceived and gestated. So there have to be parallel cases involving men. There are guys out there who took paternity tests, found out they weren’t DNA matches to their supposed children, and assumed (as you would) that their wives/girlfriends had cheated on them…and all the time they were chimeras, and the wives/girlfriends were totally faithful, and everyone’s life got fucked up for no reason.
Which I mentioned in the last paragraph of my post. But it was my understanding that, while crossing over happens sometimes, it wasn’t something that happened with every chromosome, in every generation?
Uh… again, doesn’t quite work like that for the purposes of the OP’s question about testing a family member of the putative father.
Suppose baby Adam is born and a child support case is brought against Brad. Suppose Brad is a chimera, a single individual composed of two or more cell lines derived of embryos which fused to form a single person.
Then there are two likely options. A DNA test comparison of baby Adam against a DNA sample taken from Brad could be a 1/2 match, same as a DNA test for a non-chimeric father, if the comparison sample taken from Brad is from the same part that made the sperm that gave rise to little Adam. Or the DNA could be… a 1/4 match if the comparison sample came from a part of Brad that is his “fused (full) sibling”. Either way a familial genetic relationships is proven.
But if Brad was a chimera composed of two embryos which fused… AND both of those embryos were fertilized by different fathers… then perhaps there would be no match between Adam and Brad even if Brad, the individual, really did sire Adam. But that is likely too far removed to be considered a likely source of error.
Yup. Pretty much every chromosome, in every generation. Average is between 1 and 2 per chromosome - about 33 total in every generation.
And a lack of crossovers seems to correlate with higher rates of nondisjunction leading to Trisomy 21 (Down Syndrome). In short, things tend to go wrong when crossing over does not happen.
Hm. OK, then, that’s two things I just learned. I wonder which way the causation goes, on that trisomy correlation?
From cursory reading, seems lack of crossovers causes nondisjunction.
And remember… evaluations of cross over frequencies are done in a selected population… the living. There is a fairly significant rate of chromosomal abnormalities, particularly trisomies, in spontaneous arbutuses.
So perhaps the actual average rate of cross overs in ALL conceptuses is lower than the average rate in those who live to develop to full term.
For the purposes of the OP’s question, sure, if the comparison sample comes from the part of Brad that’s a fused sibling, a familial relationship would be proven. I was thinking more about standard paternity tests, which I assumed (maybe wrongly) give a yes/no answer. So in that situation, rather than getting the result ‘Well, you’re not the father, but there’s a familial relationship, sooo…’ the father would just be given the result ‘Nope, you’re not the father.’
Again, though, I could be wrong about the way paternity test results are given.
Correlation does not imply causation.
Correlation does strongly suggest causation, but it can be in any of three different directions (A causing B, B causing A, or C causing both A and B). It’s not absolute proof (it’s always possible that it’s just a coincidence), but absolute proof doesn’t exist outside of pure math.
If that were true, most science would be impossible.
Correlation certainly suggests that causation may be involved, although it does not require it. If a correlation is found, then more research may be necessary to establish actual causation. But correlation is often an important clue that two things are related.