Two years ago (before the site change) I took the results of the ethnicity of my mom's dna and mapped them into genome mate (for her native american DNA only). I blogged about the process here. I wanted to see two things, first, how reliable was the identification of her native DNA and second, I wanted to identify if a match came from her Choctaw ancestry.
With a lot of thanks to my genealogy BFF, Sandra, who has tested numerous Choctaws for her own research, I have been able to confirm that every match that she has shared with someone of known and confirmed Choctaw ancestry (a descendant of a Dawes enrollee), has in fact been on an identified "native" segment as identified by 23andme. Segment length has been from 7-34.4 cM's with an average of 16 cM's. At this point she has around 14 matches, but we have only been able to concretely identify the ancestral family on one person.
That is not to say I question the validity at all. My mom and her father's first cousin match Sandra's Uncle, cousin and Aunt and then several more distant cousins to Sandra. Since we know of a reported relationship between the Garland family and Nahotima my ancestor, the values are reflective of the generational difference that is expected.
It is difficult to document family relationships concretely in genealogical terms among the Choctaw. That I have been successful at all has been for the grace of God that my family were among the elite half breeds and left a paper trail. The reason I may have a harder time identifying the ancestor in common with the other matches is who Sandra has tested.
While my mom has only 10 percent native DNA and my grandfather's cousin has 14 percent, Sandra has been testing folks who are at least half Choctaw/Chickasaw to Full blood as recognized by the tribe. Many of them have far from robust trees, and inevitably we come back to a full blood with a Native name at some point, and that is where I get lost. It's hard to connect a full blood to my family because we have no documentation of full families before 1830, and knowing the names of a sister or a cousin who was a full blood just isn't something we have.
Among the other things I have done though is looked at the admixture results on gedmatch for the Native folks, and there is a trend on those results and my families on the groups we see show up as having DNA on. Especially with mesoamerican and South American indian subgroups.
It would be awesome if a bunch of Choctaws decided to test DNA, and if someone would analyze the data, because I think there is so much to be learned from it.
In summary, yes the mapping of my mother's chromosomal ethnicity from 23andme appears to be accurate, since every Choctaw who has no other family relationship to another of our lines has turned out to be in fact found on one of the identified segments.