PDA

View Full Version : A curious X-match of mine



Baltimore1937
06-25-2015, 08:10 PM
I only have 13 (I think it is) X-matches after all these years. I recently got the 13th one. It is to a female in Norway (email address in Norway). Her apparent brother or close relative with the same surname also came in, but he is only an FF match (no X-match). Said female is also an FF match. That is the only basic facts I have.

My Norwegian ancestry is entirely from my father. I am assuming that this Norwegian X-match is not related to him, since I did not get a X chromosome from my dad. So this X-match must be from my mother's side. According to the family tree I constructed for myself, the only way that X-match could happen is via a line going back to Delaware Swedes in colonial times. I notice one line is given as coming from Varmland province on the Norwegian border.

AJL
07-10-2015, 07:26 PM
I notice one line is given as coming from Varmland province on the Norwegian border.

That's an interesting area because it was home to a number of Forest Finns, some of whom came to New Sweden.

Baltimore1937
07-10-2015, 08:25 PM
That's an interesting area because it was home to a number of Forest Finns, some of whom came to New Sweden.

That's what I was thinking (Forest Finns). But since I wrote the above post, I did an "in common with", and see there are also American connections over and above the Norwegian names. Actually, those names were on her brother's tree. She didn't give out any information about herself.

C J Wyatt III
07-29-2015, 02:31 PM
I only have 13 (I think it is) X-matches after all these years. I recently got the 13th one. It is to a female in Norway (email address in Norway). Her apparent brother or close relative with the same surname also came in, but he is only an FF match (no X-match). Said female is also an FF match. That is the only basic facts I have.


So you are talking about X-matches on Family Finder? Have you looked at X-matches on GEDmatch.com?

Jack

Baltimore1937
07-30-2015, 09:27 AM
So you are talking about X-matches on Family Finder? Have you looked at X-matches on GEDmatch.com?

Jack

No, I've never done anything with Gedmatch. With my wobbly WiFi connection, I don't try to do things like data downloading/uploading. I have lots of Scandinavian FF matches that I don't get involved with. Their surnames are this-son and that-dotter by the dozens.

C J Wyatt III
07-30-2015, 01:25 PM
No, I've never done anything with Gedmatch. With my wobbly WiFi connection, I don't try to do things like data downloading/uploading. I have lots of Scandinavian FF matches that I don't get involved with. Their surnames are this-son and that-dotter by the dozens.

Thank you, Baltimore1937.

Family Finder does not give you anything close to the true number of your X-DNA connections. If you were able to upload to GEDmatch, I think you would find a bunch more. My mother gets a whole lot of X-DNA matches, so I would be curious to compare. It's a long shot, but by chance do you have anyone with my surname on your FF match list?

Jack Wyatt

kjjohnston
07-30-2015, 08:39 PM
In 2014 Family Tree DNA added X chromosome matches to those who already match on the autosomes. Unfortunately some people have significant X matches with no autosomal match and will not show up as a match at all at FTDNA. The X is only one out of many chromosomes so it also takes a large population of people to get a match in the first place.

The X match list is full of people who only match at 1 cM on the X so that list of X matches is inflated, particularly for women. I am a female and I have 166 "X matches" (most are not meaningful) outside my immediate family but my father has only 4 X matches at FTDNA. He has more than a thousand autosomal matches. This gender difference is typical. My father has more total FF autosomal matches than I do and the overall numbers of females are similar to males. Endogamous populations will show proportionally more X matches than this in general.

You will often need to increase the threshold to 5+ cM or higher at Family Tree DNA to find a meaningful X match. Most of the matches for females at FTDNA are false positives and those should disappear above 10 cM.

The threshold of 1 cM causes many more matches for females than for males. Females will have a lot of pseudo-segment matches due to the fact that they have two X chromosomes and these are not phased. Males are naturally phased and the segment matches are more likely to be identical-by-descent because the sequence is not ambiguous for their single X chromosome. Therefore, the sensitivity of the test is much higher for males. The gender differences are enormous and it is not just because females have twice as many X chromosomes to match.

Kathy

C J Wyatt III
07-30-2015, 09:34 PM
Kathy, if you were looking for X-DNA matches on GEDmatch, what parameters would you use?

Jack Wyatt

dp
07-31-2015, 07:28 PM
Reading these posts got me thinking about X matches of my relatives at 23andme (Countries of Ancestry).
paternal aunt's:
http://www.anthrogenica.com/attachment.php?attachmentid=5409&d=1438370732
[the Austrian match surprised me. The aunt below is maternal 2nd cousin of the one below. The two don't share many foreign matches.]

maternal aunt's:
http://www.anthrogenica.com/attachment.php?attachmentid=5408&d=1438370722
[I was glad to see a Scandinavian match. I wonder why this one has so many more X matches?]
-dp

PS: is there an IBD/IBS threshold for X segments. Would say 12cM be enough to assume that the segment is IBD?
In other words does at least the first aunt share IBD -X- segments with foreign matches?

C J Wyatt III
08-01-2015, 01:35 AM
David,

Are your aunts' kits on GEDmatch? I'd like to take a look if it is possible.

Thanks.

Jack Wyatt

kjjohnston
08-01-2015, 03:02 PM
Kathy, if you were looking for X-DNA matches on GEDmatch, what parameters would you use?

Jack Wyatt

That is a difficult question, Jack, because male matches are so much more sensitive than female matches. I like to go above 10 cM for a female to female match and even for a male to female match. 7 cM is still a good threshold to start with. For a male to male match, a much lower cM segment can be IBD but I rarely find a common ancestor unless I also see a nice autosomal match as well.

See http://www.isogg.org/wiki/Identical_by_descent for X matching thresholds at 23andMe.

A lot of people are not looking at their X matches. There have been genealogists who were very surprised when they found a common ancestor with an X-only match - it usually is above 20 cM. Many of these are only seen at GEDmatch. However, my brother triangulates on an X-segment-only with two other males (above 20 cM) but I can't find the common ancestor. Even with good pedigrees, the ancestor is likely to be too distant in the past during American colonial times. The geographical areas could be correct though. The X pattern of inheritance does involve a lot of females and these are harder to trace. At least you can narrow the playing field with male matches because there are fewer lines to follow.

Endogamous populations (e.g. Ashkenazi) are notorious for having too many matches and pile-up regions. But we even have some success stories coming from this population.

When comparing matches, I also pay attention to the test company. You have to lower the SNP number if you are comparing V2 or V4 from 23andMe with FTDNA and AncestryDNA. Beware, because I find there are not only a lot of false negatives (because of SNP counts) but also a lot of false positives (if the cM size is too low in women). The SNP density is not great in some of these regions. I think Illumina could do a better job because there are SNPs available that are not being tested.

Always do a one-to-one comparison when evaluating the match. I play around with the parameters but have not come to any definitive conclusions about what you should or should not use. We need more real-life success stories.
Kathy

dp
08-01-2015, 04:04 PM
paternal aunt, is: M136639
maternal aunt is: M521083
they are maternal first cousins.
Yours,
dp :-)
If you match them and have a 23andme kit please send me a sharing invite. thanx.

David,

Are your aunts' kits on GEDmatch? I'd like to take a look if it is possible.

Thanks.

Jack Wyatt

khanabadoshi
08-01-2015, 04:19 PM
EDIT: I should mention this is all y-DNA. I have no X-matches, not even a list :(

It seems you guys have a good handle on matches. Can you help me understand how close of a match some is? How many cM and SNPs is significant? I am curious because my top match on gedmatch is to a Chitrali with 4-5 generations MRCA and 18cM but not enough SNP matches, 550 I think. and the next best thing is someone who seems to be Irani, with 8.5 cM and 1595 SNPs, MRCA at 6-7 generations. Exactly the generation I am looking answers for. You think I should e-mail him? If he knows the name of his great x6 grandfather that might help me out. Then again his y-DNA is J something, so it's unlikely that any fruit will bear?

Then there is a random match with someone in the Netherlands: 9.1cM/709 SNPs.
A Polish guy: 8.1cM/764SNPs

C J Wyatt III
08-01-2015, 05:33 PM
That is a difficult question, Jack, because male matches are so much more sensitive than female matches....

Kathy

Thanks Kathy for the detailed explanation. I'll take a close look at it and reply later. I have had a good bit of success using short segments so I will try to give my views about that.

Do you think we should make this discussion a thread of its own?

Jack

C J Wyatt III
08-01-2015, 06:00 PM
paternal aunt, is: M136639
maternal aunt is: M521083
they are maternal first cousins.
Yours,
dp :-)
If you match them and have a 23andme kit please send me a sharing invite. thanx.

Thanks David.

As you can tell by my reply to Kathy, my methodology is a bit out of the mainstream. That being said, I am seeing some X-DNA connections between my mother and your aunts. I'll try to put together an analysis later. In addition to my mother's autosomal results on Family Finder, we have tests pending with 23andMe and Ancestry DNA. Also she has mtDNA FMS pending. Her test will be my first exposure to 23andMe, but I certainly want to make her results accessible to anyone that is interested (that was the reason for the duplicate testing).

This might seem an odd question for a post on X-DNA matching, but I'm curious how far back with confidence that you have your paternal line. I saw something in the regular autosomal comparisons that has my curiosity.

Looking forward to continuing the discussion.

Jack

kjjohnston
08-01-2015, 06:15 PM
EDIT: I should mention this is all y-DNA. I have no X-matches, not even a list :(

It seems you guys have a good handle on matches. Can you help me understand how close of a match some is? How many cM and SNPs is significant? I am curious because my top match on gedmatch is to a Chitrali with 4-5 generations MRCA and 18cM but not enough SNP matches, 550 I think. and the next best thing is someone who seems to be Irani, with 8.5 cM and 1595 SNPs, MRCA at 6-7 generations. Exactly the generation I am looking answers for. You think I should e-mail him? If he knows the name of his great x6 grandfather that might help me out. Then again his y-DNA is J something, so it's unlikely that any fruit will bear?

Then there is a random match with someone in the Netherlands: 9.1cM/709 SNPs.
A Polish guy: 8.1cM/764SNPs

I think you meant all autosomal. You probably come from a population that is not heavily tested so you may have more difficulty finding matches. You may not get much help in pedigree building.

kjjohnston
08-01-2015, 06:22 PM
[/QUOTE]: is there an IBD/IBS threshold for X segments. Would say 12cM be enough to assume that the segment is IBD?
In other words does at least the first aunt share IBD -X- segments with foreign matches?[/QUOTE]

12 cM match between females can be IBS or so far back that you can't find a common ancestor. But you'll never know until you look. If the geographic populations don't match at all, I tend to ignore the match unless the cM size is higher than 20 or there is an autosomal match to go along with it.

kjjohnston
08-01-2015, 06:33 PM
Thanks Kathy for the detailed explanation. I'll take a close look at it and reply later. I have had a good bit of success using short segments so I will try to give my views about that.

Do you think we should make this discussion a thread of its own?

Jack

Feel free to start a new thread discussing segments or whatever you like.
Kathy

dp
08-04-2015, 06:47 PM
Between 1730's to 1830's is the century my furtherest known genealogical ancestors were born.
dp :-)

Thanks David.

As you can tell by my reply to Kathy, my methodology is a bit out of the mainstream. That being said, I am seeing some X-DNA connections between my mother and your aunts. I'll try to put together an analysis later. In addition to my mother's autosomal results on Family Finder, we have tests pending with 23andMe and Ancestry DNA. Also she has mtDNA FMS pending. Her test will be my first exposure to 23andMe, but I certainly want to make her results accessible to anyone that is interested (that was the reason for the duplicate testing).

This might seem an odd question for a post on X-DNA matching, but I'm curious how far back with confidence that you have your paternal line. I saw something in the regular autosomal comparisons that has my curiosity.

Looking forward to continuing the discussion.

Jack

khanabadoshi
08-04-2015, 06:54 PM
I think you meant all autosomal. You probably come from a population that is not heavily tested so you may have more difficulty finding matches. You may not get much help in pedigree building.

Yes, my apologies. I meant autosomal. Thanks for the correction. Yes, matches are few and far in between, that's why when I see something actually pop up on the one-on-one I take it as significant, when in fact it probably isn't.

Baltimore1937
08-04-2015, 11:52 PM
Thank you, Baltimore1937.

Family Finder does not give you anything close to the true number of your X-DNA connections. If you were able to upload to GEDmatch, I think you would find a bunch more. My mother gets a whole lot of X-DNA matches, so I would be curious to compare. It's a long shot, but by chance do you have anyone with my surname on your FF match list?

Jack Wyatt

Hi Jack,

I didn't get back to this thread for a while.

I have just one FF surnames match with "Wyatt". She gives a long list of surnames that includes Wyatt, and says all of her known origins are from the British Isles. Her own last name is not Wyatt. It looks like she may be connected to my direct maternal line in some way.

Rebecca
06-14-2016, 08:32 PM
I only have 13 (I think it is) X-matches after all these years. I recently got the 13th one. It is to a female in Norway (email address in Norway). Her apparent brother or close relative with the same surname also came in, but he is only an FF match (no X-match). Said female is also an FF match. That is the only basic facts I have.

My Norwegian ancestry is entirely from my father. I am assuming that this Norwegian X-match is not related to him, since I did not get a X chromosome from my dad. So this X-match must be from my mother's side. According to the family tree I constructed for myself, the only way that X-match could happen is via a line going back to Delaware Swedes in colonial times. I notice one line is given as coming from Varmland province on the Norwegian border.


I am tickled pink to find as many as I have found, but even more so with the family tree on family search website. I did this one to see what my DNA shows. Can anyone help me understand this?
Eurogenes EUtest V2 K15 Oracle results:

Kit A981397

Admix Results (sorted):

# Population Percent
1 North_Sea 32.59
2 Atlantic 27.94
3 Baltic 11.25
4 West_Med 10.77
5 Eastern_Euro 8.65
6 West_Asian 3.12
7 East_Med 2.89
8 Red_Sea 1.93
9 Oceanian 0.5
10 Northeast_African 0.21
11 Amerindian 0.11
12 Sub-Saharan 0.04

Single Population Sharing:

# Population (source) Distance
1 Southwest_English 3.31
2 South_Dutch 3.41
3 Southeast_English 4.07
4 North_German 5.33
5 Irish 5.66
6 North_Dutch 5.99
7 Danish 6.12
8 West_Scottish 6.27
9 West_German 6.46
10 French 7.7
11 Orcadian 7.71
12 Norwegian 9.18
13 Swedish 10.13
14 West_Norwegian 10.48
15 East_German 10.48
16 North_Swedish 12.14
17 Austrian 14.1
18 Hungarian 15.09
19 Spanish_Cataluna 15.23
20 Southwest_Finnish 15.34

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 75.3% North_Dutch + 24.7% Spanish_Cantabria @ 1.32
2 75.7% North_Dutch + 24.3% Southwest_French @ 1.33
3 76.9% North_Dutch + 23.1% Spanish_Castilla_La_Mancha @ 1.5
4 78.4% North_Dutch + 21.6% Spanish_Andalucia @ 1.53
5 82.8% Southwest_English + 17.2% Austrian @ 1.56
6 73.8% North_Dutch + 26.2% Spanish_Castilla_Y_Leon @ 1.58
7 76.8% North_Dutch + 23.2% Spanish_Valencia @ 1.59
8 77.5% North_Dutch + 22.5% Spanish_Aragon @ 1.59
9 72.4% North_Dutch + 27.6% Spanish_Cataluna @ 1.64
10 75.5% North_Dutch + 24.5% Spanish_Extremadura @ 1.66
11 87.2% Southwest_English + 12.8% Croatian @ 1.72
12 74.2% North_Dutch + 25.8% Portuguese @ 1.84
13 75.6% North_Dutch + 24.4% Spanish_Murcia @ 1.84
14 86.8% Southwest_English + 13.2% Serbian @ 1.85
15 84.5% Southwest_English + 15.5% Hungarian @ 1.85
16 51.1% Southwest_English + 48.9% South_Dutch @ 1.86
17 65.2% South_Dutch + 34.8% Irish @ 1.87
18 67.9% South_Dutch + 32.1% West_Scottish @ 1.9
19 75.3% Danish + 24.7% Spanish_Cantabria @ 1.91
20 88.6% Southwest_English + 11.4% Moldavian @ 1.91

loisrp
06-14-2016, 09:18 PM
Don't assume because your X match is in Norway that she has only Norwegian ancestry. Many Norwegians have other ancestry too. (Edit: I see the original post is from a year ago. But this still holds true.)

loisrp
06-14-2016, 09:21 PM
I am tickled pink to find as many as I have found, but even more so with the family tree on family search website. I did this one to see what my DNA shows. Can anyone help me understand this?
Eurogenes EUtest V2 K15 Oracle results:

Kit A981397


These look like the results of someone with modern northwestern European ancestry.

dp
06-15-2016, 05:48 PM
Welcome to Anthrognica!
dp :-)

I am tickled pink to find as many as I have found, but even more so with the family tree on family search website. I did this one to see what my DNA shows. Can anyone help me understand this?
Eurogenes EUtest V2 K15 Oracle results:

Kit A981397

Admix Results (sorted):

# Population Percent
1 North_Sea 32.59
2 Atlantic 27.94
3 Baltic 11.25
4 West_Med 10.77
5 Eastern_Euro 8.65
6 West_Asian 3.12
7 East_Med 2.89
8 Red_Sea 1.93
9 Oceanian 0.5
10 Northeast_African 0.21
11 Amerindian 0.11
12 Sub-Saharan 0.04

Single Population Sharing:

# Population (source) Distance
1 Southwest_English 3.31
2 South_Dutch 3.41
3 Southeast_English 4.07
4 North_German 5.33
5 Irish 5.66
6 North_Dutch 5.99
7 Danish 6.12
8 West_Scottish 6.27
9 West_German 6.46
10 French 7.7
11 Orcadian 7.71
12 Norwegian 9.18
13 Swedish 10.13
14 West_Norwegian 10.48
15 East_German 10.48
16 North_Swedish 12.14
17 Austrian 14.1
18 Hungarian 15.09
19 Spanish_Cataluna 15.23
20 Southwest_Finnish 15.34

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 75.3% North_Dutch + 24.7% Spanish_Cantabria @ 1.32
2 75.7% North_Dutch + 24.3% Southwest_French @ 1.33
3 76.9% North_Dutch + 23.1% Spanish_Castilla_La_Mancha @ 1.5
4 78.4% North_Dutch + 21.6% Spanish_Andalucia @ 1.53
5 82.8% Southwest_English + 17.2% Austrian @ 1.56
6 73.8% North_Dutch + 26.2% Spanish_Castilla_Y_Leon @ 1.58
7 76.8% North_Dutch + 23.2% Spanish_Valencia @ 1.59
8 77.5% North_Dutch + 22.5% Spanish_Aragon @ 1.59
9 72.4% North_Dutch + 27.6% Spanish_Cataluna @ 1.64
10 75.5% North_Dutch + 24.5% Spanish_Extremadura @ 1.66
11 87.2% Southwest_English + 12.8% Croatian @ 1.72
12 74.2% North_Dutch + 25.8% Portuguese @ 1.84
13 75.6% North_Dutch + 24.4% Spanish_Murcia @ 1.84
14 86.8% Southwest_English + 13.2% Serbian @ 1.85
15 84.5% Southwest_English + 15.5% Hungarian @ 1.85
16 51.1% Southwest_English + 48.9% South_Dutch @ 1.86
17 65.2% South_Dutch + 34.8% Irish @ 1.87
18 67.9% South_Dutch + 32.1% West_Scottish @ 1.9
19 75.3% Danish + 24.7% Spanish_Cantabria @ 1.91
20 88.6% Southwest_English + 11.4% Moldavian @ 1.91

vettor
06-28-2016, 03:06 AM
If I have 1% of Ashkenazi ( it was 1.5% but seems to be lowering every year ) in my X-Chr ...........how old would that be ( or number of Generations )?

Saetro
06-28-2016, 05:08 AM
If I have 1% of Ashkenazi ( it was 1.5% but seems to be lowering every year ) in my X-Chr ...........how old would that be ( or number of Generations )?

I have seen numbers all over the place with X chromosome as regards parent/child transmission and cousin matches, so presumably ethnic aspects are the same. Apart from the obvious male versus female inheritance rules, the inheritance of any segment per generation (female to anyone) can be from 0-100%.
The most plausible estimations for my matches come from GEDmatch, when you ask it to predict the linkage between matches on the X.
But that is applying averages. And it depends very much how often males are involved in the line.
I think that is one of the Tier 1 tools.

Presumably your 1% could reasonably be narrowed to something probably more than n generations ago, and maybe even suggest a specific number, but the error estimate will be huge.