Wednesday, May 1, 2013
Autosomal DNA/Family Finder Test
While attending the Ohio Genealogical Society's annual conference this past weekend, I attended a class given by Diahan Southard called Autosomal DNA for genealogists. Diahan is no longer employed by any of the companies who do DNA testing, so her information is not biased by any of the companies who are doing autosomal DNA testing. She has a website:
The main fact that Ms Southard shared that we ALL need to think about is that at this time when viewing results of autosomal testing is that 2nd cousins are approximately 99% accurate, 3rd cousins are about 90% accurate, and 4th cousins are about 50% accurate.
In my mind this means that you are very likely to have matches with people who share your gr-grandparents with an excellent accuracy. Those who share your 2-gr-grandparents are also fairly accurate. Those who share your 3-gr-grandparents are missed about half of the time. And beyond that, it is pretty hit of miss.
Another really important idea that Ms. Southard shared is that there are two reasons for a match:
1. You really DO share a common ancestor....and it is just as likely to be female as male
2. You simply share a common geographic origin. That is the people in that area were so intermarried and carried so much the same genetic material that they will all match to some extent. I interpret this to mean that I might match someone whose ancestors were living in a small village in England a century or two ago even though we can not come up with a common ancestor. We would just share much of our autosomal DNA because the families there were so intermarried that the people leaving the area many generations later would have had somewhat similar DNA.
Her information included the suggestion that a match might be because you share many small segments of DNA. That would indicate a geographic connection. With another match you might share large segments of DNA which would indicate that this person is more likely to be related to you with a common ancestor in the much closer time frame.
Now this is just my interpretation of what I heard. You can find more accurate information from more professional informants on any of the sites by the major companies and on Ms Southard's website. I wanted to share what I thought after the class because another Hawkins participant had shared his disappointment with me that he and a participant who were matches when their yDNA was compared did not show up as FF matches. I had the very same thing happen. Gene Hawkins who is my dad's closest yDNA match did not match either my dad nor my own FF. I found the above explanation of accuracy to be helpful in understanding that fact.
I took the below from the FTDNA site. It shows a comparison of FF results between three people. The gold shows the segments in common between the person doing the comparison and Anthony Bolden while the aqua shows the segments in common between the person doing the comparison and Harold Weimar. While Anthony Bolden and Harold Weimer show likely connection to the person who is doing the comparison, Mr. Bolden and Mr. Weimer are likely to be connected to the person doing the comparison by different ancestors since they do not share the connected segments with each other. On the other side of that question is the fact that I find FF matches of my own that I would expect to match each other as well as me....that is that their paper trail suggests that all three of us have the same family lines from the same area. For example all three of us have similar surnames in Eastern Kentucky in the 1800's. Yet the segments on which I match with one are not the same places as the segments that I match on the other as shown in the below example. This does not rule out our having the same common ancestors. We may have inherited different segments from our common ancestors.
Hope that I have made some sense of the question of why one might have a yDNA match but then not have a FF match with the same person.