Sunday, December 28, 2014

When a Match is not a Match?


A few days ago I was exploring my Family Finder matches at FTDNA. When I do this I first select to "Show Full View" which is just above the icon for my first match:


This allows me to see "Longest Block" and other items of interest.

I then located the person with whom I shared the longest block but had yet to find a relationship on paper. That person shares a block of 41 cMs with me on Chromosome 10 and is predicted by FTDNA to be my 2nd to 4th cousin. I then looked for others who matched the two of us. There were several who overlap part or all of this 41 cM area of Chromosome 10. Ten others overlap that shared block in amounts varying from 9.9 cMs to 29 cMs. The various relationships and common ancestors shared among the individuals in this cluster will take some time for us to try to sort out.

All of these individuals are on my dad's side of the family. In addition to matching me, they also match my paternal first cousin. Also four of them (including the one with the 41 cM match) match each of the others. All of them match at least five others in this group.  


Not all apparent matches are real.

One of the individuals in the above grouping seemed to have an exactly identical match. Both were predicted to be 5th cousins to distant cousins. In the Chromosome Browser their segments looked like this:


Upon closer examination the exactness apparently seen in the Chromosome Browser was confirmed:

Chromosome
Start
End
cMs
SNPs
10
20918456
30616944
12.34
2796
10
20918456
30616944
12.34
2756


But wait. Let's not get ahead of ourselves. It is easy to get hypnotized by the statistical precision of DNA lab reports. DNA doesn't lie and lab errors are rare. However, we must be careful to interpret the results correctly. 

My first clue that something was amiss was when I loaded these individuals into Family Finder's Matrix tool. The individual represented by the orange bar in the Chromosome Browser view of Chromosome 10 above matched me, my paternal first cousin, and seven of the others when I compared them in the Matrix tool. The apparently identical blue bar represented an individual who only matched me. Later I discovered that this second individual also matched my maternal first cousin. In other words the first (orange) matched a segment of DNA that I had inherited from my father. The second (blue) matched a segment of identical length and location that I had inherited from my mother. That double helix must be respected. The two apparent exact and identical matches with me turned out not to be matches with each other.  


4 comments:

  1. Doesn't this relate to you having two chromosome 10s, one from your mother and one from your father, not to the fact that each of those chromosomes is a double helix?

    ReplyDelete
    Replies
    1. Barbara,
      That is exactly correct. It is the contributions of the father and the contributions of the mother that taken together make up the double helix. Each chromosome by itself does not constitute a double helix.

      Delete
    2. Actually, each copy of a chromosome (maternal AND paternal) IS a double helix, made up of complementary base pairs.

      With regard to the main point of your example, I have become increasingly concerned about the possibility of false negatives at FTDNA. What if the odd man out did share that particular segment but did not meet other thresholds (e.g. the apparent requirement for segments adding up to 20 cm total)?

      Delete
    3. Thanks for confirming that, DNACousins. (I also have some potential "false negatives" amongst my relatives so would be interested in seeing that topic explored further)

      Delete