Recently, someone asked which of the major DNA testing companies test the X chromosome and which ones use the X in matching. How does this difference influence the quality of our matches?
|Vendor||X in Download File||Uses X in Matching||X Included in Total cM Count|
|Family Tree DNA||Yes||Yes (if have a match on another chromosome)||No|
*If Ancestry did utilize the X in matching, it wouldn’t benefit customers because Ancestry does not show segment information by chromosome. In other words, no chromosome browser.
Family Tree DNA includes any size X match IF and only if the two people already match on a different chromosome.
GedMatch, of course, isn’t a vendor who does DNA testing, so they don’t provide download files. They are solely on the receiving end.
X CentiMorgan Counts
Due to variations in the way vendors calculate matches and total cM counts, your mileage may vary a bit.
In other words, the 23andMe cM total, if an X match is involved, may be slightly more than a match between the same two people at Family Tree DNA, where the X match cM is not included in the cM total.
Conversely, you won’t show an X match with someone at Family Tree DNA if there isn’t also another segment on a different chromosome that matches.
In general, due to the thin spread of SNPs on the X chromosome, you will need, on average, a cM match that is twice as large as on other chromosomes to be considered of equal weight.
In other words, a 10 cM match on the X chromosome would only be genealogically equivalent to approximately a 5 cM match on any other chromosome.
X matches really can’t be evaluated by the same rules as other chromosomes due both to their SNP paucity and their inheritance path, which is why most vendors don’t include those segments in the total cM count.
While including the X chromosome cM count is problematic, X matching can be a huge benefit because of the unique inheritance path of the X chromosome.
In the article, X Marks the Spot, we discussed the inheritance path of the X chromosome for both males and females. Females inherit an X chromosome from both father and mother, which recombines just like chromosomes 1-22. However, men only inherit an X from their mother, because they inherit a Y from their father instead of the X. Therefore, males will only inherit an X from their mother, and females will only inherit their father’s mother’s X chromosome.
Charting Companion software works with your genealogy software of choice to produce a lovely fan chart where the contributors of my X chromosome are charted in color, above. You can read more about Charting Companion here.
The great news is that if you and a match share a significant portion of the X chromosome, meaning more than 15 cM which reduces the likelihood of an identical by chance match, the common ancestor (on that segment) has to come from an ancestor in your direct X path.
I’m always excited to see with whom I share an X. That piece of information alone helps me focus my ancestor detective efforts on a specific portion of my tree.
Some X segments can remain intact for generations and may be very old. So don’t be surprised if the common ancestor of the X segment and another matching segment may not be the same ancestor.
Sorting by X
I wasn’t able to find a way to sort by X chromosome matches at 23andMe, but you can sort by the X at both Family Tree DNA and GedMatch.
At GedMatch, X matching shows on the one-to-many match page. You can sort by either Total X cM or Largest X cM by using the up and down arrows, at right, below, in the X DNA columns.
After you identify an X match, be sure to run the X one-to-one match option to verify.
My GedMatch matches cause me to wonder if 23andMe is using a different reporting threshold for the X chromosome, because one of my matches at GedMatch is a close family member with no X match at 23andMe, but a total of 32 X cM and with a longest segment of 14 X cM at GedMatch.
That same individual matches me with the largest X segment of 14 cM at Family Tree DNA as well.
Family Tree DNA X Match Phasing
At Family Tree DNA, on your Family Finder matches page, just click on the X-Match header (at right, below) to bring all of your X matches to the top of your list.
If you have linked any kits of relatives to your tree, you will see numbers of phased kits on the maternal and paternal tabs with the red and blue male and female icons. In the example above, I have 3313 matches total, with 744 being paternal, 586 being maternal.
Next, click on the maternal or paternal tab to see only the people with X matches who match you on the your maternal and paternal lines. Matches are automatically sorted into maternal and paternal “buckets” for you. Remember to check the size of the X match before deciding about relevance.
Who is your largest X match that you don’t already know? Maybe you can find your common ancestor today.
I receive a small contribution when you click on some of the links to vendors in my articles. This does NOT increase the price you pay but helps me to keep the lights on and this informational blog free for everyone. Please click on the links in the articles or to the vendors below if you are purchasing products or DNA testing.
Thank you so much.
DNA Purchases and Free Transfers
- Family Tree DNA
- MyHeritage DNA only
- MyHeritage DNA plus Health
- MyHeritage FREE DNA file upload
- 23andMe Ancestry
- 23andMe Ancestry Plus Health
- Legacy Tree Genealogists for genealogy research
We had a adoptee contact our family who had a inherited family heart condition. Using his last name at birth we managed to trace that unlucky inheritance back to my great grandmother, however it was my grandfather’s mother who carried that heart cond. My grandfather, my mother and at least 3 other siblings all died young due to sudden heart failure. I have 10X Dna match and I am 89 cM with the adoptee (male). My maternal 1st cousin has 56X match but only has 12 cM. The adoptee is in my gr grandmother’s brother’s lineage. He comes down through 3 generations of males and lastly his mother. Other DNA matches confirm the family is correct. I am not sure how that can happen due to the male to male inheritance. Any insights? His heart issues sound exactly like my mother’s, her fathers, her brothers (2 of them) and sister.
I’d have to draw this out and I can’t do that right now, but remember that a 10 cM X match is only like a 5 cM regular match and maybe noise.
You won’t have male to male inheritance on the X chromosome, but there is another possibility. I have a 3rd cousin with whom I share several autosomal segments that all appear to have come from his father’s side. However, I also share two fairly large segments that can only have come from his mother’s side.
The solution? It turns out that his parents were cousins. In this case, actually rather close cousins. (1st cousins, if I recall correctly.) But even if they had only been distant cousins, this still could be the explanation.
Remember, segments on the X chromosomes are often handed down for more generations than autosomal segments of the same size. It depends very much on the combination of sexes involved in the transmission.
For example, I share two segments on the X chromosome with a likely 3rd cousin once removed. These are fairly large segments — 25.7 cM and 40.7 cM. I believe I’ve been able to identify the likely source of these segments as Rachel Jones, wife of Henry “Hunting Shirt” Sizemore.
I did this by identify the possible X-donors in each generation until I was able to find one common to both of us. Since we’re both male, our fathers can immediately be eliminated as possibilities, along with all of their ancestors.
In addition, I know for reasons I won’t go into at the moment that all but 20 cM of my X chromosome came from my maternal grandfather. Both matching segments were from completely outside this region, which means they came from my grandfather’s mother.
Going up from my great grandmother’s father, there did not seem to be any connection to my 3rd cousin once removed. Her mother was a different story. My great grandmother’s mother was a Sizemore — any of whom theoretically could have transmitted one or both segments were: Dillion Blevins Asher, Sarah Collett, Henry “Hunting Shirt” Sizemore … and Rachel Jones. Of these, the only one with an X-path to my cousin was Rachel Jones.
Further, the X-path from Rachel to my cousin alternated sexes in every generation — which is the X path with the fewest possible recombinations. On my side, the path included just one female to female transmission, and the rest were either female to male or male to female.
However, as I noted at the beginning, my cousin’s parents were related to each other. I can also trace a path from his *father* to common ancestors, including even Rachel Jones. It’s just that no X chromosome DNA could have traveled along such a path to my cousin. But certainly autosomal DNA could have.
What I’m trying to say is that common ancestors can sometimes have more than one path to a given descendant or descendants. It so happens that there are far fewer possible X paths, but unless you know that the heart condition actually relates to the X chromosome itself, you don’t necessarily need an X path on both sides — other paths might also exist.
At FTDNA they will not recognize my mother as my mother (hence giving me two “both” matches. I match my sister and her children on my paternal side only (according to their diagram). We are identified as sisters sharing 2555cM. We are all linked in my and in her tree. I match her as “both” on her page. The chromosome browser show us sharing the whole x, I have contacted them multiple times and they tell me it is because my mother had no siblings. That has to be bogus because my sister and I match on her page. And she had no siblings for my sister either. This is driving me crazy
Did you and your mother both test at Family Tree DNA, or is one of your kits a transfer? That seems odd.
The problem is that some Ancestry kits are downloading with too few locations tested. Here’s a fix for the problem. https://dna-explained.com/2017/10/27/ftdna-unlock-sale-upload-fix-triangulation/
Thank-you for this valuable information on the X Chromosome and usefulness in matching on some of the top DNA products. Could you also comment on the usefulness of the Y-search tools/ database that I believe is supported by Family Tree DNA. Can a female use this database to find male matches using autosomal or maternal DNA?
No, YSearch is just for males who upload their Y DNA results. It was created at a time when there were several vendors, but now that there is only one, it’s much less necessary.
GEDmatch has helped me make sense in mapping my mom’s chaotic X by giving long segments to work with. 30% of her 4000+ matches at Family Finder are identified as X matches, but most are with tiny segments. Less than 20 are over 10 cM and they haven’t been particularly helpful. At GEDmatch I was able to search on X matches over 15 cM from the “One to Many” search function on the home page by clicking X and changing the cM. 200 some matches were listed to gather matching segments for the map. While it’s still a work in progress, I’ve been able to clearly identify clusters and have triangulated at least three segments to “X possible” ancestors. I’ve also been to create my X from her as well as from Dad. Interesting that all my X matches over 15 cM at GEDmatch also X match either of my parents. That’s what I’d expect but it doesn’t seem to work that way with all those little scraps at Family Finder.
Dear Roberta, I love reading your blog and have especially enjpyed reading about your incredible to trip to Ireland. Thank you. I am a 77-year old adoptee and, while I met my birth mother later in her life, she died over 10 years ago of metastatic breast cancer to the brain without revealing the identity of my birth father. Around 4 years ago, I had atDNA testing and learned that my birth father was of Finnish descent with a small amount of Volga-Ural Russian. I also learned that I had around 3,000 Finnish DNA cousins living in Finland. I thought that it would be easy to identify my birth father through DNA testing. Wrong! Due to endogamy, a single Finnish DNA cousin match and I could share DNA with 300 – 600 other cousin matches. I have very few Finnish DNA cousin matches residing in the US because my Finnish grandparents were recent immigrants – probably 1890 – 1910. My inquiry to you is about the X chromosome. One of my Finnish DNA cousin matches is a woman and we share 31.3 total cM and 15.8 longest cM. This is the largest amount of X cM of any other Finnish DNA cousin match, both in amount and length. Does this mean that my Finnish grandmother and my female Finnish DNA cousin match were related through a common ancestor? She’s a 3.9 cousin match on Gedmatch and I’ve communicated with her son, who manages her account, and he’s very receptive to helping me. How far back would I need to trace to identify a potential common ancestor and which branches of her family tree would I follow? My highest Finnish DNA cousin is a 3rd cousin on AncestryDNA. He’s an anonymous male, with a male administrator and no family tree. I’ve sent the administrator email messages – 5 months apart – and he’s never responded. I know he’s seen the message because the only times he’s logged on is to view my messages. Thank you, again, and I send blessings to you for all that you do to further DNA research. You give me hope. 🙂
It could be. Unfortunately, the X chromosome is less than reliable. My suggestion would be to contact Jennifer Zinck who specializes in this type of family reconstruction. Click on the Help tab at the top of the blog for her link and also that of Legacy Tree who I also refer to regularly.
Thank you for your promptand helpful response.
I just love all this! Thanks Roberta! Not only because I’m learning a lot, but because it’s a challenge: I’ll work on it as time permits just for fun!! These are my DNA cousins too!! I was inspired to look again at my gedmatch.com X matches, where I had a few months ago, when using my phased paternal kit, found a profile who only matches me on my X chromosome. I had written the person, we shared trees, but we each agreed that we couldn’t figure it out. Yet it was a really big match, to me! Like 50 cM! Well, today I looked at that match again, and there was a second profile with (almost) the exact same overlap! And, they do match each other on X, and only X! Since I saw it was a 23andMe kit number I went there, and lo and behold, there was that profile, and yet another one there with the almost exact same match numbers! (Except some of the SNP numbers were pretty off, though, so what you said, Roberta, about 23andMe and gedmatch.com having different ways of measuring is probably true). So that’s four of us who all match, all different surnames and the overlap is 49.57 cM according to 23andMe, and 50.1 cM according to gedmatch. (the number of SNPs seems to fluctuate so I do not know what’s going on with that, the numbers are slightly different when I use my non-phased paternal kit, as they will be I know.) Also, the two profiles who match me on X on gedmatch, their X match with each other is only 20 or so more cM in genetic distance than my own with either of them (73.8 cM as opposed to 50.1 cM). Because it’s so much bigger than any others except for my mom and sister, Isn’t it likely Identical by descent if it’s 50cM? Not noise? I hope so! Even if it’s 10 generations back…If I find these ancestors, it will be like finding the ancient artifacts!!
That’s fascinating and 50cM is quite large. It’s like 25 cM on other chromosomes and the changes of that being noise are pretty remote. So I can’t tell you positively that it IS ancestral, but I’d lean that way heavily. It’s fairly compelling since you have a phased kit too, so it phases for at least one generation. I know you are dying to know who contributed that big chunk of DNA:)
Question, does a person need to match you on autosomal DNA also to be a viable X match? I have several X matches greater than 30 cM who have no match on autosomal/
No, they don’t.
I have around forty matches who match me at 30cm or larger (largest segment). One is 49.1 and ten others over forty. Most do not match me autosomally. This is certainly worth a look. Thank you.
I wanted to add this. The results for X on One to Many are different on the Tier 1 ‘One to Many’. Why would that be?
I don’t know. If they are significantly different, you might want to let GedMatch know.
I just looked at the fan chart and I also have Claxton’s in my family line. 🙂
Thank you so much for this article, Roberta. It came at the right time in my research for me to understand the differing programs I’m managing and how to sort through all my own as well as my father and siblings matches to understand how to identify which ones are in my direct line. Very much appreciated.
I was pleased to see your blog on the X chromosome as I am having a great deal of difficulty understanding the inheritance thereof. I understand that the Y climbs the male line back to the proverbial Adam. I also understand that, being female, one of my X’s climbs the female line back to the proverbial Eve and that my other X came from my father who received it intact from his mother. Here’s where it gets hairy for me. You and other knowledgeable people have stated that a male inherits his mother’s X but does not pass it on. Did my father not pass his mother’s X on to me? Did my mother not pass her mother’s X on to me? Isn’t that how I received two X’s? My daughter received one X from me and one from her father. The X that my daughter received from me would have to be the one that I got from my mother or the maternal mitochondrial line would be broken. How it it determined that she gets that particular X? It seems to me that I have two mitochondrial lines. I think I have a second one going from my paternal grandmother back.
I know my reasoning on this is faulty but I am not able to determine where the fallacy lies. I expect to have a “duh” moment when my errors are pointed out to me.
The father doesn’t pass the X to HIS SON. He does to his daughters. This article might help. https://dna-explained.com/2012/10/01/4-kinds-of-dna-for-genetic-genealogy/ As well as this one: https://dna-explained.com/2017/07/26/x-matching-and-mitochondrial-dna-is-not-the-same-thing/
Both articles were most helpful and I now see where I went down the wrong path. Thank you so much.
I was surprised to see 23&Me’s ethnicity map for the X. Gedmatch doesn’t even do that (or am I missing something?) The X I got from my mother shows way more Amerindian than expected although I guess it’s hard to tell with just a visual and no cM data. Looks 75-80% and I’m only 10%. If ethnicity can really be “concentrated” in the X it might be helpful in finding small amounts missed by autosomal.
No GedMatch doesn’t use the X in ethnicity either. X matching would at least allow you to focus on a particular path.
I suppose it isn’t so much that an ethnicity may become “more concentrated” on the X chromosome as that it might be “better preserved”.
Whether this happens depends on a number of factors, but one is simply how many opportunities there are for the chromosome to be split up. The more the chromosome is split and recombined, the greater the chance that a particular segment will either be made smaller or disappear altogether.
Now, because a mother has two X chromosomes, these may recombine with one another in the process of creating an ovum, and therefore the X chromosome she passes on to a son or daughter may include DNA from both of her parents. Or it may include DNA from just one parent — either one.
A son, however, will always pass on a nearly exact copy of the X chromosome he inherited from his mother, *if* he has a daughter. If he has a son, that son will inherit a copy of the father’s Y chromosome and not his X chromosome.
So in a mother-to-child transmission of an X chromosome, the X *may* be a recombination that includes segments from both of her parents, though it can be from just one. In a father-daughter transmission, the X will be the same as the one he inherited from his mother.
A line with alternating sexes will only have had half as many “opportunities” for recombination as a line of all females. (And of course a line of all males will not pass on an X chromosome at all.)
Effectively, that means that a line that runs female-male-female-male-female-male-female will have had no more opportunities for recombination than a line that runs female-female-female-female(or male). For both of these, there will have been three times that crossovers could have occurred on the X chromosome.
Also, keep in mind that in any of the mother-to-child transmissions, it’s also possible for no apparent recombination to have occurred. I mentioned in another comment that two of my siblings — one brother and one sister — inherited a maternal X chromosome containing DNA from our mother’s *father* only.
My brother also has a daughter, and her paternal X chromosome is of course a match for her father’s only X, which is a match for our grandfather’s only X, so that *all* of my niece’s paternal X chromosome came from her 2nd great grandmother. If that 2nd great grandmother had been fully Native American instead of only part, then my niece would have an X chromosome that was 100% Native, even though my niece’s autosomes would only show a fraction as much.
I should also have noted that even if my great grandmother — my niece’s 2nd great grandmother — had been fully Native American (instead of a quarter or less), that wouldn’t necessarily mean that 23andMe’s Ancestry Composition would necessarily reflect that.
There are two technical concepts that apply to the test: one is precision, which essentially is a measure of the degree to which a reported ancestry is actually correct; and the other is recall, which is a measure of how well an ancestry that is present is actually detected.
23andMe says that for “Native American”, precision is 99% and recall is 86%. This is based on test runs involving samples of “known” ancestry. In other words, 99% of the time that the test identified a given segment as Native American, it really was. But, only about 86% of the time were Native American segments actually detected as such.
This doesn’t mean the segments were instead identified as European, but more likely they were identified as either generic “Native American and East Asian” or as “East Asian” — at least *most* of the time, even when a more precise identification was not made.
Thank you Roberta and Gary for taking the time to provide such helpful and interesting information. It seems the X has come mostly intact from my maternal line great-grandparents since my father and maternal grandfather were fully European.
I get emotional to see this bold signature of ancestors whose world was vanquished but not their DNA. Now I will have to find out if this X has recombined in my sons and grandchildren. Good thing there’s a sale!
This blog post has been very helpful to me. I’m a man (which has some real advantages when dealing with X-DNA) and I am fortunate to have inherited a very hardy string of X-DNA that I can trace to my 5th great grandparents. I have used this X-DNA segment to identify MANY relatives at 23andMe and at Gedmatch. It’s particularly frustrating that Family Tree DNA (FTDNA), which I have always regarded as the premier testing site for genetic genealogy, treats the X-chromosome so shabbily. I have 10 kits at FTDNA, and I don’t think I’ve EVER found a relative there using X-DNA. I have ONE kit at 23andMe, and it has yielded at least 20 confirmed matches. I have over 100 X-DNA matches at Gedmatch with whom I share segments over 10cM, and 33 with shared segments over 20cM. I wonder how many matches I have at FTDNA that are inaccessible because of the way it treats the X chromosome. My situation may be unique (but I don’t think it’s terribly unique), but FTDNA’s treatment of the X chromosome makes me wonder whether I’ve tested at the wrong company for all of these years…
What would you make of a 59 cM X-dna match for my cousin and a 10 cM x-dna match for me to a recent adoptee who contacted us. We have traced him to our maternal grandfather’s mother who has the same last name as his birth mother (Fuller). We have determined through other matches he descends from our gr grandmother’s brother and in fact down through 3 generations of fathers with same family name. We identified a grandfather (Fuller) we believe is his maternal grandfather, however he would have had to father his unknown mother at age 16 if adoption nonidentifying info is correct with her age of 17 when said adoptee was born. We contacted Fuller DNA match family members who heard rumors that the possible grandfather might have married or fathered a child, but no birth records, marriage records either can be found. Yet we have this X-dna hanging out there. Interesting I match the adoptee at 102 cM’s and I have the lower X-dna match. My cousin only matches him at 11 cMs and she has the higher X-match. Our matches are from GEDmatch by the way. Do you have any insight on this situation?
I would suggest that you contact http://www.dnaadoption.com who works with these kinds of things regularly:)
ok i am confused. i wanted to take one of these tests because my father never met his biological father. my father is deceased now. my sister and i thought by us taking the test we would get an idea of what our biological grandfather’s ancestry was. but reading the article that wont happen because we dont have a y chromosome (and no brother). is this am accurate assumption to make? thanks so much
This just applies to the X chromosome, not the rest. You’re fine. Contact http://www.dnaadoption.com for help.
Even though neither you nor your sister will have inherited your father’s Y chromosome or his father’s X chromosome, it’s still a virtual certainty that you inherited a lot of autosomal DNA from your grandfather.
In fact, most people inherit somewhere around 25% of their DNA from each grandparent. Since you and your sister are unlikely to have inherited the *same* 25%, for the two of you together the percentage will likely be significantly higher.
But, one of the problems you’ll face is trying to determine just what DNA came from your paternal grandfather. To do that, you’ll first have to figure out what is from your father.
How do you do this? You didn’t say whether you’re paternal half sisters or full sisters. If it’s the latter, you can *generally* presume that anywhere you match your sister is on your father’s side. However, many people find that the more relatives they discover, they more they begin to see at least a few folks who may be related on both sides — even if their parents are not related, or not closely related.
If you’re full sisters, things get a bit complicated since you’ll have many matches with your sister on each side. In fact, in some cases you’ll match your sister on *both* sides. However, at some companies you will have no way of knowing this.
Ancestry, for example, does not offer any sort of chromosome browser. You can *partially* get around this by looking at relatives both you and a match have in common, as long as your parents don’t share many relatives themselves. If they’re both from the same area, this will not always be the case — even when they’re not closely related.
Also, while FTDNA *does* offer a chromosome browser, they don’t distinguish between half identical matching (matching on just one side in a given region) and completely identical matching (matching on both sides in a given region).
So, if you’re full sisters, you really should also test your mother — if that’s possible. If not, it can still be useful to test the closest relatives you can.
I was lucky enough to discover that four of my paternal 1st cousins once removed have tested at Ancestry. They’re all great grandsons of my father’s parents. *Generally*, any of my Ancestry relatives who has one of these cousins as a shared relative is likely to be related on my father’s side.
I also have five paternal relatives in the 3rd cousin category, though one is actually a 2nd cousin once removed and another is a 3rd cousin once removed. Three of these are on my paternal grandfather’s side, and two are on my paternal grandmother’s side.
The point is, you and your sister are likely to share a great deal of DNA with your father’s father, and therefore have a possibility of finding paternal relatives. The trick is going to be determining when a given relative is paternal, if you and your sister are full siblings and cannot also test your mother. It may still be possible, but require a bit more time and effort.
I have a question. You stated that “females will only inherit their father’s mother’s X chromosome.” I am wondering if I have learned something in error, as I thought that females inherited portions of their father’s and their mother’s X, and therefore their paternal grandmother’s, and their maternal grandfather and grandmother?
Reading back over that section again, I understand what you were trying to explain, that females inherit their father’s mother’s X only (paternal grandfather)only, and not her father’s father’s X (paternal grandfather).
Yes, from their father, they will only inherit their father’s mother’s X. They also inherit a copy of the X from their mother, which can be part of both of her parents’ X or one or the other.
Roberta do you wish you’d never posted this lol? I think I’m pretty smart: I was a pre-med major in college (haha I never went to med school–smart me I wanted to have a life–no offense doctors… too bad I became a lawyer 😉 I had my own fruit flies family in my lab with all of the different colors of eyes, and I was supposed to be able to predict what percentage of their offspring would have what colors of eyes, which I did, and I would go count them based on their eye color, in my Senior level Genetics class at my university, but imagine my surprise when I had no idea how X transmission worked until Roberta Estes explained it! And it’s still hard to understand! I have a hard time not laughing when others think they know how it works but they don’t!! It’s elusive, to say the least!! 🙂
I have one distant female cousin, double 5th, who shows a lot of cM, but varies considerably across the sites. Mostly due to how X is counted (or if it’s counted).
3 segments – Chr 7, X1 and x2
(I have triangulated the segment X1, to a woman, Magdalena Voorhees, b1739. A Dutch NY person. Via seven or 8 X matches through 4 of her children)
chr 7 – 15.1 cM
Total 41 autosomal only
Rank 542 out of 2009, 3rd to 5th
Chr 7 17.7 cM
Total 1.02%, abt 76 cM autosomal + X
Rank #2 out of 1374, 3rd to 4th
GedMatch (kit from FTDNA)
Chr 7 18.8 cM
Total 18.8 autosomal only
Rank about 500 out of 2000, 4.7 gen (3rd to 4th)
Total @ 1 cM (31 segments) like FTDNA (14 segments)
NOT a match. Therefore <6 cM autosomal/phased. No X counted
Included in a DNA Circle.
Was a match, pre-Timber.
So the Big Picture is similar except at AncestryDNA. Since 23andMe does count X, my rank pops up to #2 there.
If I were to nitpick, there are a lot of differences in the lengths and totals. More than I like. For instance, 23andMe is higher on X1 but lower on X2. And GedMatch picks up a lot more of the tiny segments than FTDNA, even though the data is from FTDNA.
Although I did not include START and STOP points, they were extremely similar. Slightly different for 23andMe.
23andMe and FTDNA calculate the lengths of *all* of the chromosomes differently. That includes the X chromosome. By FTDNA’s calculation, the full length of the X chromosome is 195.93 cM; by 23andMe’s calculation, it’s only 182 cM. (Source: https://isogg.org/wiki/CentiMorgan)
GEDmatch has different numbers from both different from both, although the number is almost the same as FTDNA’s on the X chromosome. (They use an integer, 196 cM.) In the case of other chromosomes, their number is closer to that of 23andMe. In fact, for chromosome 7, GEDmatch and 23andMe both use 187 cM, while FTDNA uses 180.79 cM. (Source as above.)
Why this difference? Consider that the centimorgan does not represent an absolute length, the way that Mb does. (It’s 1,000 base pairs.) Instead, it’s based on the probability of there being a 1% chance of a crossover occurring in a specific (real) length of DNA in a single recombination event.
That’s while you can’t really say how many base pairs there are in a cM. You’ll sometimes see “1M” or one million given, but that’s perhaps an *average*. In reality, one stretch of DNA might have fewer base pairs in a cM, while another might have more — which is why you can’t really do a visual comparison of matches which are in different locations.
And there’s an important factor that is often overlooked. Not only is the recombination rate different in different parts of each chromosome, it’s also significantly different between males and females. In fact, it’s higher in women than in men by perhaps 50% or more.
So what does this mean for the calculation of cM for our matches? Well, obviously in order to give a number, the company — whether 23andMe, FTDNA, or GEDmatch — has to have some sort of estimated recombination rate for the region in question. The problem is, right off the bat the average should be different for different combinations of sexes in that line.
Take your surname line, for example. If there is no NPE between you and, say, your 5th great grandfather, any segment you inherited from this great grandfather can be expected to be larger than segments you inherited from, say, your 5th great grandmother on your mtDNA line. On the other hand, with more though smaller segments being passed on, you may have a greater chance of inheriting at least *something*.
Of course, for X chromosomes you have the fact that the only time they really *can* recombine is when passed on by mothers. Fathers will only pass on their single X to daughters, and then without any meaningful recombination — because there’s no other X for his X to recombine with. (The two regions, PAR1 and PAR2, are both quite small and add a layer of complication we don’t need to go into.)
What it means, though, is that over time you’re more likely to inherit significant lengths of DNA on your X chromosome when there are more males on the line between you and the contributor. (Actually, everyone in between in a contributor, but I mean the “original” for that particular segment.)
I’ll add that some pretty impressive-sized X chromosome segments can be passed along quite a distance. The niece I mentioned — my brother’s daughter — essentially inherited her entire paternal X chromosome from just one 2nd great grandmother, Hannah (Muncy) Babcock. We’re talking about 182 cM or 193.93 cM, depending on whether you use 23andMe’s number or FTDNA’s number.
That’s because the X chromosome my brother inherited from our mother contains no contribution from our mother’s mother — that is, it contains DNA only from our maternal grandfather. So it’s all from his mother Hannah.
Further, although the X chromosome Hannah passed on to my grandfather likely contained some DNA from each of her parents, that means that *at least* half of my niece’s paternal X must be from either our grandfather’s maternal grandfather or his maternal grandmother. So my niece inherited not less than 91 cM from a single 3rd great grandmother or 3rd great grandfather! (And more than that for one of them, if she didn’t inherit an equal amount from both of them.)
What does it mean when I was told by CRI Genetics that I don’t have enough X chromosomes to do an mDna test?
The X and mitochondrial are entirely different tests. It means you need to test with a different company. Use Family Tree DNA. They also provide matching for you with others. Here’s the link. http://www.kqzyfj.com/click-6754800-13710356
Thanks for posting this… and VERY TIMELY for adopted paternal 1895 grandmother born in Philadelphia.. a-parents married in 1891, shows up as 1 live/ 2 as “Dau” in 1900. In 1910 relationship rubbed out/rough paper written over- Dau?? with “ad dau—??” A-parents both d. before turned 21 and had court appointed guardians.
Two Ancestry(V1) matches since 2015 ( father-dau) at 203/40 cMs 2nd-4th cousin… not logged in since original test results w/ no tree and no reply to A.com messages. Surrogate Match is by son of 1895 adoptee.. On Feb 1, new Ancestry(V2) match 183cMs 3rd cousin pops up with tree stub to a father & mother and grandmother/father. A pvt/mirror built up from stub nets the 2015 kits as a half uncle to the Feb new match. Request to upload to Chr browser – is being “thought about” by half uncles family….
In mid March 23andMe (V5?) results gave up two more kits of 138 and 40cMs (V??) the larger includes 38cMs X match ( Hg is not a match). Now have 2nd and 3rd matches from each of three children by two husbands to the new Feb 1 match’s grandmother…. and a cluster of first-cousin -all more distant matches to surrogate son- to a very hi frequency surname but their roots are Phila/NJ and could very well be up tree collateral to grt grt gnm of same surname – hopefully proving spousal origin for Surrogate to common grand parents of the proximate cluster. Have been “painting” and now have three large segments for X off Gedmatch, 23andME and 3 doz- 9-12 cMs X-Matches- majority are inclusive of 3 Lg Xsegs..
I’ve been trying to verify if Ancestry used X chromo in match counts- especially the 203 cMs whose daughter is 40 cMs- makes huge difference in interpreting without a chromo browser.. Thanks again.
I’m a bloke, so I got my X from my mum. Of my 1200 or so FTDNA matches, only a few match on the X. Three of these are definitely on my paternal side – FTDNA says so, and they also match with cousins and uncle/aunts on that side. So they are paternal all right. All three share almost exactly the same segments with me, according to the chromosome browser. It’s not much – about a couple of centimorgans.
So, is this likely to be a chance match, or does it indicate that my Mum’s ancestors were somehow related to my Dad’s in some past time?
That’s too small to consider. An X segment needs to be about twice as big as a segment in another chromosome to be as valid.
A couple of years ago, I was looking at my small X matches to see if I could at least use them for phasing parents.
At that time, I had 11 X matches at FTDNA, who did not have matches to my mother (who is also at FTDNA but not my father). I was wondering why my apparently paternal autosomal matches also showed maternal X matches (albeit small).
Out of the 11, three of them also had uploaded to GedMatch. Since the X tools are more extensive there, I was able to see that on all three, the little X segments did match little X segments on my mother’s file.
So its very possible to have a decent autosomal paternal match that also contains very small X maternal matching segments. The fact that I had paternal matching segments did not stop me from accruing 1-3 cM maternal X segments.
Both of my parents were Colonial. Not related as far as I know. Maybe very distantly.
I, too, have noticed some small X matches with relatives on my father’s side. In some cases, closer examination has suggested that a match may not be “real”, but in a few cases it’s looked as if I might be related to the person on both sides.
This hasn’t always meant I share DNA on both sides, only that the individuals seem to be related both to relatives of my mother and relatives of my father. This has been a real puzzle to me, since not only do my parents not seem to be related in the least, but they came from two different parts of the country — my father from Pennsylvania, and my mother from Mississippi.
But recently, something occurred to me that might explain how a person might independently be related to me on both parents’ sides. My mother’s hometown in Mississippi was Biloxi, which happens to be the location of Keesler Air Force Base — which is a major training facility.
Many besides my father — including, probably, cousins of his — would have gone there for basic training. Since my mother’s roots in Biloxi go back to the very beginning of the 1700s and include many people recognized as “founders”, one or more of these cousins might have ended up marrying a local girl (or guy) who happened to be a cousin of my mother.
The point is, though there seems to be no real overlap between my father’s ancestry and my mother’s — at least not later than immigration — there is a possible mechanism for a connection. (Of course, there are other possibilities as well, but the Keesler connection is one which would ensure people from all over the country ending up at least for a while in my mother’s home area, even with no family from there or anywhere near there).
While 23andMe does not allow you to search your x-matches directly, you can obtain a list of your x-matches if you download your aggregate data. Once you have the names of your x-matches, you can search for them via the DNA Relatives tool. I went through every x-match listed on my aggregate data Excel worksheet. When viewing their profiles, I made sure to utilize the ‘notes’ tab, typing “X-match” on each profile. The handy thing about using the notes tab is that anything you post there is searchable. Now I can simply type “X-match” in the search box and a list of all of my x-matches will be generated. Very handy!
Roberta thank you for this great blog about X – Just what i was looking for ! I am looking at AncestryDNA and trying to understand the Totalcm values, and if what we see at the i dot info details is including X cm or not. After reading this post i still have a question. If there were a cousin match with me showing 24 cM i think these are the possible outcomes am i correct? 1. 1 segment at 24 cM such like a longest block or 2. 2 segments both at 12 cM each OR 2 segments approx one at 6 cM and one at 18 cM. – 3. 2 segments, one seg at 24cM and the other segment a x and no value is added to the total cM given. I am just wondering about the AncestryDNA cm values and number of segments are arruarte or a little off due to a few having a x segment in the number count.
Ancestry does not include the X, and it strips out some DNA that it considers too matchy, so their numbers are always off.
Amongst my X-only matches at 23and me there are 5 large matches that are all in the same area of the X chromosome ranging in size fro 54 to 40 cm. This seems to be a sizeable chunk, yet we seem to be distant relations.
What can I reasonably learn from these matches?
My best guess is this chunk is from a common late 18th century Pennsylvania German ancestor :
Me (Male)>Mother >Mother> Father> Mother> Father
Does that sound reasonable?
It really is unfortunate that FTDNA does not provide access to X-only matches. It would be interesting to see if if I had similar matches there.
Yes, I would say that’s reasonable. Perhaps additional matches will shed light on the common ancestor.
It sounds quite reasonable to me, too. I’m sure you’ve noticed that out of five transmissions of X chromosome DNA on your side, two of these were from father to daughter. The father’s X chromosome would have been passed on to his daughter unchanged. So in reality there were only three times that recombination *could* take place.
Given a starting length of the X chromosome of about 195 cM (https://isogg.org/wiki/CentiMorgan) and only three subsequent recombinations, it isn’t that difficult to see how at least 54 cM of your 3rd great grandfather’s X chromosome could have been passed down to you.
And you might actually have inherited a much longer segment than this from him. What you’re seeing with your matches is just the amount of *overlap*, since your matches didn’t necessarily inherit exactly the same segment leads to the same possible contributor.
I might also note that if any of your matches are male, and if both of you have daughters, then this same matching on the X chromosome would extend for one more generation on both sides. It *might* also extend for another generation with a female match if she has a child of either sex, but further recombination would also be a possibility on this side.
Thanks, Gary. This is very helpful. Interestingly, this is the one chomosome that 23andme identifies as French and German at the speculative level. I thought all of my German ancestry had been overwhelmed by my Irish ancestry. On Ancestry my mother and brother are both assigned Pennsylvania Settler ancestry, but I was not. In my case this must be because Ancestry excludes the X–the only place I have significant Pennsylvania ancestry.
I have requested this change in a future version and I am hopeful that the change will be made.
That would be wonderful, Roberta!
The only reason I can see for excluding X chromosome matches that are not accompanied by any autosomal matches is the because — relative to their size — it’s possible for X chromosome matches to be more distant.
This is true, but they can also be easier to trace. For males, one entire side can immediately be excluded. For females, that would be one out of three grandparents. Even more ancestors can be excluded from search *if* you can narrow down who the various regions of your X chromosome actually came from, as opposed to who they *might* have come from.
For example, as I’ve mentioned in a previous reply, I have a brother whose entire X chromosome is from our maternal grandfather. For him, *any* match on the X chromosome says which of our eight great grandparents he can begin his search with.
My X chromosome matches my brother’s except for 20 cM surrounding the centromere. So we match for about 89%. That means if I have a match on the X chromosome, without knowing where it is I can say it’s about 8 times more likely to be on my grandfather’s side than on my grandmother’s side.
For this reason, *Ancestry* could make up — a little bit — for its lack of a chromosome browser if they’d add one more piece of information to what they’re already telling us. Currently, Ancestry says how many cM we share with each match, in how many segments. But as I understand it, this is limited to autosomal sharing.
If Ancestry would simply add X chromosome sharing (when it is present), that would be very useful for many of us. For males, it would allow those trying to figure out how they’re related to focus their search from the outset on just one side; for females, on just three of four grandparents.
To some extent, “Shared Matches” can help in zeroing in on possible common ancestry, but this is more dependent on how many shared matches there are; and on whether we know the side(s) of these shared matches. In some instances, shared matches might be on *both* sides, in which case knowing whether any shared DNA includes the X chromosome could really be useful.
For a long time little was understood about the X and its value. It has about half the SNPs of other chromosomes. It also doesn’t divide as often. However, we know more now, so I hope the matching is updated in the future. However, it means that the matching has to be rerun.
Thank you, Roberta! Much appreciated.
Roberta, just this past week I was contacted by an individual that said her Uncle and I were matched at FTDNA in Family Finder. She also stated that we were an X-match. When I looked at the matches page for my FTDNA Family Finder, her Uncle and I were autosomal matches with 51 cM shared, longest block 25, but there was no X-match listed. Last year I had ported my Ancestry results to FTDNA just to compare results of both my tests. When I looked at the results from the Ancestry data, I found that her Uncle and I were autosomal matches with 61 cM shared, longest block 25, and there was an X-match. I am an X-match to both my brother and my son, but her Uncle only has an autosomal match with me and my brother on our FTDNA samples. Her Uncle and my brother were autosomal matches with 46 cM shared, longest block 25. He does not match my son at all on Family Finder. Using the chromosome browser to compare her Uncle with my DNA data from Ancestry (ported to FTDNA,) the X-match shows only at the 1 cM level and goes away if I change it to 5 cMs. It states that we match on Chromosome X at the genomic position 62191204 – 68327715 with 2.90 cMs shared and 291 SNPs. Using the chromosome browser to compare my sample done at FTDNA there is still no X-match even if the setting is at 1cM. FTDNA predicts that her Uncle, my brother and I are 2nd – 4th cousins in relationship.
She does not recognize any of the surnames I have listed for my direct maternal and paternal lines. I did recognize one of her surnames, which is her Uncle’s surname. That surname is only collaterally related in my maternal line. I have not researched this collateral line beyond the fact that my maternal great-grandfather’s sister married a man with the same surname as her Uncle. The sister had three sons. I am lost in trying to find an explanation for this. Is this a match by chance? Why would my ported Ancestry sample data show a match but FTDNA sample data not show a match?
Regarding the X match. It’s very small. All it would take is a single no-call in the wrong place to push it below the threshold. That’s the only thing I can figure. It’s too small to be significant.
I am intrigued by X chromosome inheritance and results on 23andme v Ancestry. I’ve got almost my entire immediate family uploaded to gedmatch for comparison. One of my sons tested through 23andme, one through Ancestry. In gedmatch, the 23andme son matches a mixture of me, his siblings, my brother, maternal grandparents and my father, so his X was the mixture of my two X’s. The Ancestry son ONLY matches me, his siblings, my father, and others through my fathers side. e.g. Ancestry son matches my father on X per gedmatch at 196cm, 23andme son matches my father at 48.6cm…. Could Ancestry son have inherited an entirely clean one sided X? Is that possible? Or is there a difference in testing companies that can cause this?
Yes, happens often with the X.
Pingback: 4 Kinds of DNA for Genetic Genealogy | DNAeXplained – Genetic Genealogy
Pingback: Triangulation Resources In One Place | DNAeXplained – Genetic Genealogy
If two females share 181.84 cms on the X chromosome, (23 and me test) would they be half sisters? They are getting some inconsistent results when looking at the daughter of one of the two potential half sister who is indicated as a first cousin once removed.
Thanks in advance for your help!
The X chromosome does not have a role in detecting sibling ship. Look at full vs half identical regions. I wrote about that.
Well, the X chromosome cannot by itself determine whether two people are siblings, or what kind of siblings if they are siblings. Yet it can certainly be suggestive. For example, if you already have reason to believe two people are half siblings, it can sometimes tell you whether they’re maternal or paternal.
For example, if you know that two people are a half brother and a half sister (or if they’re half brothers), if they share a significant portion of the X chromosome then they must be maternally related. A man gets his X chromosome from his mother only.
I also have an example of two women who share an entire X chromosome but are definitely *not* sisters. My #1 sister shares 100% of her maternal X chromosome with one of our nieces. She’s the daughter of my #1 brother.
The reason they share an entire X chromosome is that my #1 brother and #1 sister share an entire X — which happens to be my brother’s only copy and my sister’s maternal copyl We’ve been able to determine that it is entirely from our mother’s father, and consequently from his mother — my siblings’ and my great grandmother, and my niece’s 2nd great grandmother.
In addition, I match this niece everywhere on the X chromosome *except* for a 20 cM region surrounding the centromere. This 20 cM came from my maternal grandmother. My daughter matches her cousin in the same way I do — that is, everywhere but in that 20 cM region.
Still, I think if two women — or two men, or a man and a women — share an entire X chromosome, it’s very likely to indicate a rather close relationship. The X chromosome typically does recombine when it can, but not always.
(Obviously, my mother passed on an unrecombined copy to at least two of her six children. And of course fathers always pass on an unrecombined copy.)
Hello. I have a question. For the 23 chromosome (X), do I use the total amount of shared X DNA with a match or only look at the size of the largest segment on the X Chromosome when looking into X matches. For example, I have a match on Ged Match that reports 16cM as the X DNA match but family tree DNA shows 25cM X match. Upon closer inspection I saw that there are 2 segments on the X Chromosome shared between me and that match. The one is 16cM and the other is 9cM. Ged Match just reported the largest one while family tree reported the total amount of shared cM’s on the X Chromosome.
Which should I look at, largest segment or total shared cM’s on the X Chromosome?
I usually use the total. I paint my result to track by relative.