Strange matching with three persons

[jonr]

Three of my matches on my X-chromosome overlap on positions 71991396 - 78395137. I'm a male, so I only have one X-chromosome and I got it from my mother. They are all female so they got two X-chromosomes. This means that at least two of these matches share this part of the X-chromosome on the same chromosome and would match each other. These two matches and I probably all share the same common ancestor somewhere on my mother's side.

I was surprised when I checked the Family Finder matrix, none of them match each other, they only match me. How is this possible? Where does my logic fail?

Thanks,
Jon

[prairielad]

What is the SNP total and cM length?

it is possible that they do not share enough autosomal DNA (chromosome 1 thru 22) for FTDNA to match them to one another. Currently the X chromosome is not used in matching, it is only shown if you happen to match someone on chromosomes 1 thru 22.

The other reason may be due to false match, Identical by State, between you and each of the women if SNP and cM length is low. Alogrithm maybe picking values from each of the woman's 2 X's and creating a match to you.

Only way to know for sure,if they do not match one another, is to compare woman's raw data via David Pike's Utilities, or X one to one on Gedmatch if they also utilise site.

[jonr]

Oh OK, I didn't know the X chromosome was excluded in the matching algorithm, that of course explains it!

Thanks a lot.

[JOlson]

Triangulation

Just because A matches B and A matches C, doesn't mean that B also matches C. This may help explain:

Triangulation

.

[jonr]

Just because A matches B and A matches C, doesn't mean that B also matches C. This may help explain:

Triangulation

.

Yeah, but in my case we're taking about shared segments of DNA, not just any match.

[JOlson]

Yeah, but in my case we're taking about shared segments of DNA, not just any match.

Yes, I'm talking about matching segments. Consider this simplified example:

Person A:
A A A A
C C C C

Person B:
C C C C
C C C C

Person C:
A A A A
A A A A

Person A matches B, and Person A matches C, but Person B does not match person C.

[jonr]

Yes, I'm talking about matching segments. Consider this simplified example:

Person A:
A A A A
C C C C

Person B:
C C C C
C C C C

Person C:
A A A A
A A A A

Person A matches B, and Person A matches C, but Person B does not match person C.

Your simplified example doesn't take in account that there are four of us, plus we're talking about the X chromosome, I only have one of those. Perhaps I'm being ignorant, but I'd say at least two of "the others" and I descend from a common ancestor.

[awheaton]

prairielad answered the original posters question very well.

Although we don't have the cM value, the listed start and end is 71991396 - 78395137 and the poster is male and only matches females on the X. From that I would assume the cM value is very low and identical by state.

I am male and only have female X matches. For comparison I have the following X match segments with similar locations to the posters and these are completely IBS and so are not genealogically relevant.

Start - End - cM - SNP's
70587634 - 78517168 - 1.49 - 525
70211882 - 77800225 - 1.71 - 500
70587634 - 78769541 - 1.5 - 550

[jonr]

prairielad answered the original posters question very well.

Yes, I was satisfied with prairielad's answer.

Although we don't have the cM value, the listed start and end is 71991396 - 78395137 and the poster is male and only matches females on the X. From that I would assume the cM value is very low and identical by state.

Yeah, here they are:
A: Me
B: 70587634 to 78769541 for a total of 1.5 cM.
C: 70587634 to 78395137 for a total of 1.48 cM.
D: 71991396 to 80020571 for a total of 1.01 cM.

B & C also overlap on some smaller fragments on 1 thru 22, but not enough to be considered matches.

[awheaton]

Definitely IBS segments due to common ancestral origins so unfortunately not useful.

[Kathy Johnston]

Yes, I was satisfied with prairielad's answer.


Yeah, here they are:
A: Me
B: 70587634 to 78769541 for a total of 1.5 cM.
C: 70587634 to 78395137 for a total of 1.48 cM.
D: 71991396 to 80020571 for a total of 1.01 cM.

B & C also overlap on some smaller fragments on 1 thru 22, but not enough to be considered matches.

For those of you who have these 1 cM matches in this block, can you check your raw data if you are a male and see which haplotype block you match here:
http://1drv.ms/1eykBQY

You will have to convert from 23andMe Version 2 to FTDNA. Is your raw data build 36 or 37? The position numbers may not be exactly the same unless you have build 36. You can always check the rs (reference sequence) numbers. The same strands (orientation) will also be needed.

Many people are getting IBS matches at exactly the same spot but I would like to know if they are all the same haplotypes. I suspect most of you are plain vanilla N. European around position 75 million, but I could be wrong. It would be nice if you all had the same downstream SNPs. Anybody can let me know off-list if you prefer.

[hansonrf]

Rutgers says:

Map Interpolator Results

No. Chr Query_Posn Female_cM
1 23 71991396 96.5498
2 23 78395137 97.4779

0.9281 cM

c'mon...

[Kathy Johnston]

Rutgers says:

Map Interpolator Results

No. Chr Query_Posn Female_cM
1 23 71991396 96.5498
2 23 78395137 97.4779

0.9281 cM

c'mon...

1 cM is usually ignored because one is unlikely to find a match within a genealogical time period, but two males can be IBD at this level according to work done at 23andMe. This particular location on the X (70,000,000 to 80,000,000) is much more likely to show a match between males so I was wondering what the raw data showed. I think this is one of those common ancient biogeographical regions that needs to be looked at more closely. Personally, I find these regions to be informative for BGA even if you don't expect to have pedigree matches. So I choose not to ignore the raw data.