Is this even possible? Complete X-chromosome match for distant cousin.

+7 votes
Unusual match on FamilyTreeDNA. I tested on Ancestry and uploaded to FTDNA. Looking at "X-Matches". Most of them are just a few cM. About a dozen have between 10 cM and 40 cM.

There's one match with 196 cM on the X-chromosome (that's the entire chromosome). For autosomal DNA the longest segment is 8 cM, nothing else longer than 4 cM.

So, matching the entire X-Chromosome would be expected for some Immediate Family members. And I understand that, under unusual circumstances, an entire X-chromosome can be passed down a couple of generations. But with only 8 cM of autosomal matching, this is probably a 4th or 5th cousin.

Have you ever heard of anything like this?

Computer error???
WikiTree profile: Robert Cline
in Genealogy Help by Robert Cline G2G3 (4k points)
Interesting discovery, Robert, and below your post are numerous posts about x-chromosome matches and DNA and triangulations and other points to consider.

2 Answers

+11 votes
Best answer
There's been some discussion of this over on the Facebook Genetic Genealogy group recently.  If this was someone who uploaded a test from another site to FTDNA, it's very likely that it's an artifact.  I had the same thing pop up recently on my FTDNA account.  Apparently some sites (like MyHeritage) don't report X data.  When that kit gets uploaded to FTDNA, the blank data gets interpreted as a complete match for some reason.  In my case, I found the match over on GEDMatch and did a one-to-one comparison, and there was no X match.
by Lisa Hazard G2G6 Pilot (122k points)
selected by Robert Cline
That's correct. FTDNA looks for contradictions to a match (opposite homozygotes, e.g. AA in one party and GG in the other party). If there is no data, there are no contradictions.

I'd mark this, with Ann's comment, as best answer, but in this case I think that's for Robert to do, if the explanation makes sense to him.

But I do think the result--a full 196cM match--is almost certainly false. One interesting addendum is that I had previously uploaded my AncestryDNA data to FTDNA, and decided to buy a new Family Finder test during the last sale just to see their results with their new GSA chip. The Family Finder atDNA test returned 2.28% no-calls, which is very definitely on the high side.

I have a single X-chromosome. The AncestryDNA and the FTDNA tests both looked at exactly the same chromosome. But using the FTDNA Chromosome Browser to compare the two kits, with a threshold of 7cM set, here's what the graphic looks like:

There are actually 7 segments of 7cM or greater shown in the detailed segment data, but only 5 shown in the "chromosome view"; omitted are segments equalling 12.11cM and 22.05cM. I have no explanation why the graphic view disagrees with the detailed view.

Lacking even more of an explanation, to me, is that the comparison equals a total of 7 segments, 11,575 SNPs compared (the X is about 155.2 million base pairs in length), and 138.63cM. The difference in microarray chipsets I understand, but I don't think that's enough to explain the loss of 57cM--about 30% of the chromosome--when comparing two identical haploid chromosomes.

So Robert has what looks to be a relatively distant cousin with whom FTDNA shows him as sharing 100% of the X-chromosome, and I have two identical X-chromosomes shown as sharing only 70.7%.

It might be worthwhile, in the future, digging a bit to see if my unusual result is a one-off oddity, or if there is a repeatable pattern of anomalous data being reported.

Oh; a quick relevant but not pertinent aside. When the testing companies report X-chromosome data for males, the raw data show that the entire chromosome is homozygous. Males only receive a haploid X--meaning one of a pair of chromosomes (diploid). Since the basic reporting formats from our genotyping microarray testing companies show a pair of alleles at each reference sequence, in the report they simply duplicate the single allele at each X-chromosome position. In other words, a male's X will always look like: AA, GG, AA, CC, TT and so on, rather than heterozygous alleles, one nucleotide on one arm of the diploid chromosome, and a different value on the other arm.

My FTDNA GSA vs Ancestry v1 upload looks similar, with several segments on the X chromosome. However, the FTDNA GSA kit looks fine when compared to my 23andMe v4 (custom chip) and v5 (GSA chip) transfers. Go figure!
FTDNA needs to learn that absence of evidence is not evidence of absence.  (I think this is the third person I've explained this error to since I learned about it myself just a month or so ago.  It's not a rare thing.)
Speaking as a guy who's no stranger to number-crunching and algorithms, what they're doing is just trashy. Amateur Hour stuff.

The sad truth is that usually people who come up with this kind of stuff don't even have the good sense to be embarrassed when it goes horribly wrong. They just make excuses for it - it's beyond their comprehension that what they did was simply of low quality, because they know no other way. The FTDNA management probably doesn't understand that their guys screwed up, either. So it may take practically forever to see it fixed.
+1 vote
I think the entire chromosome gets passed down from mother to son. So it'd be interesting to find out if there are a fair number of males in the lines going back to the MACRs (although obviously not two in a row).

My maternal grandfather has his mother's X, my mother has to have the whole thing, and that has a 50/50 chance of being passed to me. It turns out that it apparently DIDN'T get passed to me, since my only known X match goes back to MACRS up my maternal line, but that's besides the point - I had a 50/50 chance of getting gt- grandma's X.

8cm of auDNA matching can be as close as a 3C. Really, I've even seen a 2C1R that low, but it's pretty uncommon.
by Frank Stanley G2G6 Mach 6 (63.1k points)
The X chromosome from mother to son would be expected to recombine, but an unchanged transfer of one of the pair of Xs is certainly possible. Father to daughter is an exact transfer as dad has no other X to recombine with.

Thanks Tim! I presume you are right. I only ever looked into this topic once, for somebody else, and undoubtedly could stand to review it.

Last I checked, I only had one X match, on GEDmatch, and I didn't really need it to figure out how we were related. So I don't seem to have much use for it, with my own test results.

Sounds like others have this figured out, the answer being both surprising and disturbing. What else are they screwing up, "under the hood", that we don't know about?

Related questions

+11 votes
2 answers
646 views asked Oct 28, 2017 in Policy and Style by William Foster G2G6 Mach 8 (88.1k points)
+11 votes
1 answer
250 views asked May 29, 2016 in The Tree House by Dave Dardinger G2G6 Pilot (406k points)
+7 votes
3 answers
293 views asked Jan 5, 2014 in Policy and Style by Veronica Williams G2G6 Pilot (136k points)
+8 votes
2 answers
+3 votes
2 answers
+12 votes
3 answers
196 views asked Apr 5, 2019 in WikiTree Tech by Peter Roberts G2G6 Pilot (524k points)
+5 votes
3 answers
+6 votes
2 answers
245 views asked Mar 24, 2018 in The Tree House by Chris Hampson G2G6 Pilot (104k points)
+25 votes
2 answers

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright