Thursday, 2 January 2014

Family Finder X-chromosome matching

Family Tree DNA have announced the release of a new X-chromosome matching feature as part of their Family Finder test. This is a particularly exciting new feature because of the very special inheritance pattern of the X-chromosome.  Autosomal DNA is inherited in a random fashion from all of our ancestors and consequently there is no easy way to determine which line you match someone on without doing all the hard work of comparing family trees and mapping the individual segments to specific chromosomes. The ability to view X-chromosome matches will allow us to refine a subset of our matches to a more limited number of lineages. For information on the inheritance of the X-chromosome see the article on X-chromosome testing in the ISOGG Wiki. The Wiki article also contains links to a number of very useful downloadable charts where you can add in the details of the relevant ancestors.

I've had a brief play with the new X-chromosome feature and I think that FTDNA have done a very nice of job integrating these new tools into the Family Finder test. This is a long promised and most welcome new feature.

My mum currently has 185 Family Finder matches but only 41 of these people (including me!) also match her on the X-chromosome. My dad has 235 matches but only three of these matches (again including me!) also match him on the X-chromosome. I I have 203 Family Finder matches. Strangely I have 59 X-chromosome matches, 15 more matches than my mum and dad's combined X-chromosome matches. When I previously compared my match list with my mum and dad's matches I found that around 18% of my matches did not appear on my parents' match lists. This situation arises because some segments are what are known as compound segments. For an explanation of this phenomenon see this FTDNA FAQ. Presumably a similar phenomenon has occurred with my X-chromosome matches.

The following screenshot shows the chromosome browser comparison for my father and his three X-chromosome matches (click on the image to enlarge it). You can clearly see the X-chromosome that I have inherited from my dad.

The matching X-chromosome segment data can be downloaded from the Chromosome Browser view into a spreadsheet.  The X-chromosome is listed in the spreadsheet as chromosome 98. It turns out that my dad's other two X-chromosome matches both have very small matching segments sharing just 1.01 and 1.21 centiMorgans respectively. They are both predicted to be quite distant relations - a fourth to distant cousin and a fifth to distant cousin. Not surprisingly, as the segments are so small, neither of these two matches appear in my own match list. These are both likely to be very distant matches and it will probably not be possible to find a genealogical connection. However, you can see that it could potentially be very exciting if there were closer matches with much larger shared segments.

The following walkthrough of the new X-chromosome matching service has kindly been provided by Family Tree DNA. Dummy accounts have been used for the screenshots. Click on the images to enlarge them.

Here is the menu link to the Family Finder Matches menu.

On the Family Finder Matches page, there is a filter to show only X-matches.

When you use the X-match filter on a male Family Finder kit, you only get matches from the maternal X-chromosome.

Next, like other Family Finder matches you can expand the advanced bar for a match and click to add the match to the Compare in Chromosome Browser list.

Matches are added to the Compare in Chromosome Browser list. You could go right to the Chromosome Browser by clicking on the compare arrow at this point.

Next we can also go right to the Chromosome Browser.

The Chromosome Browser also lets you filter the match list by X-matches.

Here are three immediate relatives. The first two share X-chromosome DNA. The third (green) one does not.

When we scroll down to the X at the bottom, we see that X-matching is displayed for the first two but not the third.

Moving to the Advanced Matching page, X-chromosome matches have also been integrated.

X-match is an option that can be checked alongside other types of testing


Maurice Gleeson said...

Thanks Debbie. Nice post. I completely agree that the excess matches you have over your mum and dad are likely to be false positive matches -"compound segments" as FTDNA call them. Ann Turner has written a very good article about this phenomenon in Jogg 2 years ago (). It would be interesting to know what percentage of segments are false positives for a given segment length, similar to the analysis that John Walden has done for autosomal DNA inheritance. The additional interesting factor about the X is that there are likely to be fewer false positives when comparing two males, slightly more when comparing a male and female, and most when comparing two females. This occurs because males only have a single X chromosome for comparison and can therefore be considered "pre-phased". It may be that 23andme have already accumulated a lot of data to explore this more comprehensively.

Maurice Gleeson said...

And here's Ann's article -

Debbie Kennett said...

Thanks Maurice. I'd forgotten about Ann's article so thanks for reminding me.

It seems that females are much more likely to get false positive X-chromosome matches. 23andMe apparently adopt different thresholds depending on the gender of the match:

X (male vs male): 200 SNPs, 1 cM
X (male vs female): 600 SNPs, 6 cM
X (female vs female): 1200 SNPs, 6 cM

See Kathy Johnston's post in the FTDNA Forums:

A lot of the X-chromosome matches seem to be showing up near the centromere and it may be that there's something odd about this segment. See Ann Turner's post in the FTDNA Forums:

It's going to be fun making sense of all this! It's certainly helps testing lots of other close family members.