24 October 2016

Who I am: DNA testing and relatedness

This is one in a series of posts on the search for my biological family.

In this post, I will look at DNA testing as it results to matching me to other people who have also done DNA tests. I will assume you know the basics of how DNA works. The more DNA you match, the closer you are to that person. Conversely, less DNA means a more distant relationship. And unless you know specifically the relationship that a person shares with you, it is difficult to determine that from DNA alone.

First, consider this table as a general idea of how much shared DNA means in terms of a relationship:

Approximate amount of shared DNA (in centimorgans)Possible relationship
3,475Parent, child, or identical twin
2,400—2,800Full sibling (including fraternal twins)
1,450—2,050Grandparent, aunt, uncle, half—sibling
680—1,1501st cousin, great—grandparent
200—6202nd cousin
90—1803rd cousin
20—854th cousin
6—20Distant cousin: 5th cousin — 8th cousins

Let's take a look at my results from Ancestry.com. My top match is a woman who shares 455 centimorgans shared across 22 DNA segments. Ancestry characterizes  this relationship as a 1st-2nd cousin. If I knew the specific relationship that we share, I could confirm the accuracy--but I don't. And most unfortunately, this woman was also adopted. She does a little about her birth family, but not enough to be helpful to me. Because she is about a generation older than me, I suspect that we are first cousins once removed--that is--one of my great grandparents is also her grandparent. Since neither of us know who that might be, it's not a lot to go on.

I have a lot of other matches on Ancestry--147 total, in fact. However, only a handful are even characterized as 3rd cousins. Most are 4th cousins or more distant than that.

On FamilyTreeDNA, my closest match shares only 127 centimorgans. I contacted this man, who just happens to be a chemistry professor at the University of Maryland. We've been in contact for a while now, and he's been very interested in helping me. This is great. However, he too knows that we have little to go on. So we're still moving very slowly.  FamilyTreeDNA also matches me with over 1,000 people--many more than Ancestry--but again, most of these are fairly distant matches. There's always a chance that someone knows something, so I do try to contact my top matches. But most people don't know much beyond their first cousins.

Lastly, GEDmatch.com shows over 1,000 matches that share at least 10 centimorgans. Here again, though, the closest match has a genetic distance of 4.0. Not terribly helpful.

You might be thinking--can't I generate a massive family tree using all of these people and find where they match? In theory, yes. In reality, it's not that simple. Most genealogy sites publicly share family trees, but limit what most users can see to those members who have died. As a result, the most useful matches are generally people who are alive not available on those trees. Some people will publicly post more, but it's really hit or miss.

In a future post, I'll talk about the second DNA test I took, the Y-DNA test.
Post a Comment