I not only was one of the first people to undergo testing for genetic genealogy, I also probably among the most tested.
I have been tested by Family Tree DNA, Oxford Ancestors, the old Relative Genetics, Sorenson Molecular Genealogy Foundation, DNATribes, EthnoAncestry, Trace Genetics, DNAAncestry, DNA Print Genomics. I hope haven't forgotten anybody.
One of the more shocking things to me was I found to be in three haplogroups. FTDNA found I was in the G group, a rare one but one linked with Ashkenazi Jews. That made sense.
Oxford found I was in the R group with the Vikings. Relative Genetic found I was an I1a, among the Norsemen.
Viking. Norseman. Jew. I'm sure my Lithuanian and Latvian ancestors encountered people from Scandinavia back when.But these results did not resonate. (Wikipedia explains the terms Viking and Norseman at en.wikipedia.org/wiki/Norsemen)
It turns out that genetic companies typically predict your haplogroup rather than testing specifically for specific SNP markers.
I asked Relative Genetics and a year ago they ran a SNP they found I indeed was a G. Specifically, they found I was G1. But again that seemed wrong. The G SNP, EthnoAncestry and FTDNA eventually all put me in the newly discovered G5 group, a sort of lost tribe of undetermined origin primarily with other Ashkenazi Jews from the Lithuanian hood.
Oxford did not respond to my queries.
Maybe most people may not run into this problem because the predictions work for them. On the other hand, most people are probably not tested by multiple organizations so they don't know the difference.
Scott Woodward, director of Sorenson Molecular Genealogy Foundation, while in my Chicago office recently.told me such discrepancies are not unheard of. He said he has a staffer who has received different results from each genetics companies.
Consumers have a hard enough time understanding these results without coping with this issue.
And New York Times reported Nov. 25 "DNA Tests Find Branches But Few Roots (www.nytimes.com/2007/11/25/business/25dna.html?ref=business) about African Americans who have received conflicting results. Henry Louis Gates Jr., who helped popularize DNA testing and now is involved in AfricanDNA, a collaboration with Family Tree DNA, told the Times about how he received conflicting results: one company informed him that his maternal ancestry went back to Egypt, with a Nubian ethnic group, while another company told the Harvard prof later that his maternal roots were European.
Likewise, CBS' 60 Minutes recently reported on a similar case when a woman was tested by three different companies.
(www.cbsnews.com/stories/2007/10/05/60minutes/main3334427.shtml)
Stanford professor Hank Greely told 60 Minutes he thinks the problem is hype, rather than fraud: "You know, beer commercials imply that drinking their beer will
make beautiful women fall all over you. I think the genetic genealogy
companies don't go below the normal standards of the marketplace. But
they don't go above it either. Some do a better job than
others, but there's not one that couldn't improve. And that bothers me
because they're using science to sell their product. And science is
about the whole truth."
The genetic companies need to do a better job of explaining what the results mean and what they don't mean. Consumers deserve no less. Eventually, the haplo-hype could come to be considered fraud.
Howard:
From my vantage point as the G Project Administrator I see the
converse of this frequently. A member joins as a predicted (red) G or
maybe just an ambiguous (hyphen). They either order a DeepSNP-G test
or FTDNA does a SNP test per their Quality Assurance Program
http://www.familytreedna.com/SNP_assurance.html
Suddenly, I have an E or an I or some other haplogroup in our project!
My process is to remove them, but first to enroll them in the correct
project for their haplogroup. The E and I Project Administrators and
I have become good friends ... I do recruiting for them! ;-)
Regards,
Peter T. Christy, Administrator
Haplogroup G Project
http://www.familytreedna.com/public/G-YDNA
Posted by: Peter T. Christy | November 26, 2007 at 12:53 PM