Update on 1/12/2015: I am again responding to a new blog post written by Robert Estes on 1/5/2015. Her post Anzick Matching Update is her admission that she used an old Gedmatch kit number for her Clovis Anzick research protocol. In this post, she again reiterates her methodology as well as her justification for including the M haplogroup on her list all the while ignoring the facts. The facts are below. She has also failed to answer any of my questions so I have also listed them below.
1) It is unwise to compare mtDNA haplogroup assignments to autosomal DNA. The main reason is that we are talking about two different types of DNA. Extrapolating mtDNA info from living people who have multiple ethnic admixtures and then comparing to an ancient Native American sample is seriously flawed. One can, in fact, match an ancient Native American, like Clovis, and have a non-Native American DNA. Roberta somehow conveniently misses the fact that on the F999919 Clovis Gedmatch One-To-Many list, there were a lot of people who matched Clovis who did not have a Native American haplogroup. Why did she not include these matches on her New Native American Haplogroup list? Was it because they are well-known non-Native American haplogroups? Is her continued inclusion of the some of M haplogroup subclades her attempt to discover something new? Why does she assume that just because someone matches a Native American autosomally that this means that they automatically have a Native American mtDNA?
2) The 2007 article she references has been brought into question by several known genetic genealogists, who are also experts in mtDNA analysis, like Ann Turner, Ugo Pereto, James Lick, Claudio Bravi and others. The M sample was not fully sequenced and was more likely to be an X haplogroup upon further analysis. I may not be an expert in genetic genealogy, but I certainly reached out to some of the best before I even wrote my posts. How come Roberta has not commented on their responses which I have reported in my blog posts?
3) Her continued inclusion of M subclade haplogroups on her Native American Haplogroup list, all the while maintaining that M has not been proven to be Native American, is very misleading and disingenuous because it gives people the false impression that they are Native American haplogroups. Lay readers will just look at the headline and make that assumption without reading the small print. Why is Roberta ignoring the fact that her posts are misleading? Is it really enough to justify the inclusion of M haplogroup, without a shred of evidence, just because she can? Should one even publish “research notes” that are not based on current data/facts?
These are questions my inquiring mind would like to know. However, Roberta has failed to answer any of my legitimate questions for months so I don’t see her changing her modus operandi today. This is what has led me to write my posts in the first place.
Update on 12/24/2014: I felt the need to share this with my readers. I was just made aware of the fact that Roberta Estes admitted today that she ran her initial results using one of the older Clovis Anzick Gedmatch kits numbers. As a result, the methodology she used calls into question her whole research protocol. I am taking her admission below that what I wrote in this blog post is correct.
A continuation of my previous blog post…
I did not address the other M subclades that were mentioned in Roberta Estes “New Native American Mitochondrial Haplogroup” list in my last blog post because it focused specifically on M23. This blog post, however, seeks to do just that because I really don’t want the public to be misled into thinking that these other M subclades are, in fact, Native American as well. Given the stature that Roberta Estes has in the genetic genealogy community, I would really like to see her to remove these M subclades from her working hypothesis. The facts just don’t add up to them being Native American at all.
I tried to replicate how Roberta Estes did her research to come up with her inclusion of the M haplogroup subclades in her hypothesis. I went back and looked at all the Clovis matches to see what M haplogroup subclades showed up and if they matched the ones Roberta mentioned on her blog (M1a, M1a1e, M1b1, M23, M3, M30c, M51, M5b3e, M7b1’2, M9a3a/M9a1ac1a). In order to accomplish this, I repeated the steps Roberta used to make her hypothesis. This meant bringing up the One-to-Many matches of the last known Gedmatch kit number we have for Clovis Anzick (F999919) and looking at the mtDNAs of the Clovis matches. When I reduced the cM level to 1 cM, I was able to pull up 1500 matches. However, only one had a M subclade of M7b1’2. I did not find any indication of any of the others she mentioned.
I can only assume that Roberta used a previous version of the Clovis DNA profile. I know there were several as I matched the first Clovis DNA profile on Gedmatch, sharing 20 cMs with 7 cM as the largest segment, but did not match later Clovis DNA profiles. If this is in fact the case, than I believe Roberta should have taken this new Clovis match info into account when she updated her list on 12/7/2014.
As I mentioned in my previous post, the macro-haplogroup M covers a wide geographical area.
I looked up the geographical locations of each M subclade Roberta mentioned and found that they were not Native American or were out of the timeframe to be relevant for any comparison to Clovis Anzick. For example, M9a1a1c1a (formerly M9a3a), though geographical close– if you consider Siberia– to being Native American is dated by Behar to be 4221.4 years +/- 3456 old and therefore is nowhere near the 13,000-15,000 age range of Clovis.
The M subclades Roberta mentioned on her blog cover the following geographical areas:
M1a, M1a1e, M1b1 -North Africa, East African, and the Middle East
M23 – Only Madagascar
M3- Southeast Asia
M30c- South Asia
M51- Indonesia, Cambodia, Vietnam, Nepal, and Laos
M5b3e- I could not locate this subclade so it may be an error
M7b1’2- East Asia
M9a3a ( now known as M9a1a1c1a)- Japan, Siberia, Tibet, China, and Mongolia.
The M Haplgroup and Native Americans
I received some great feedback on my last blog post. One was from Ian Logan, another mtDNA expert, who likewise confirmed the Malagasy origins of the M23 haplogroup. Perhaps, one of the most telling comments that I received came from Dr. Ann Turner, a well-known genetic genealogy pioneer and mtDNA expert herself. I had included, in my blog post, the 2007 article titled “Mitochondrial Haplogroup M Discovered in Prehistoric Native Americans” by Ripan Malhi, et. al. This article is the one that Roberta cites in order to include M subclades as Native American in her hypothesis. Dr. Turner made the following comments:
Dr. Ann Turner also consulted with Dr. Ugo Perego, another expert on Native American mtDNA, about the results of the 2007 M sample. Dr. Perego stated:
“Unfortunately, we might never know the true answer [because the remains have been reburied], but I am with you in thinking that it was probably a false positive for M and most likely an X, which would have been still quite interesting as ancient X’s are not that common.”
Both Dr. Turner and Dr. Perego believe that the M sample the article was based on most likely tested false positive for M when it was probably an X sample. Again, the M sample was never fully sequenced as Roberta herself acknowledges.
In conclusion, I have no idea how Roberta came up with her “New Native American Mitochondrial Haplogroups” list. Several people have asked her for an explanation of her methodology to no avail. I could not duplicate Roberta’s methodology as much as I tried. Because of this and the fact that the M sample she refers to was probably an X, I would like to see her remove the M subclades from her list of “New Native American Mitochondrial Haplogroups.” The evidence is simply not there to make that claim that they are “Native American Mitochondrial Haplogroups.”
We need to remember that individuals today have complex, multiple ancestries that may not be reflected in their mtDNA. I am one such person. I consider myself culturally African-American and Puerto Rican and my ethnic admixture is tri-racial (46% Sub-Saharan African, 46% European, and 8% Native American). But, when you look at my H1ag1 mtDNA, it is European. I am a perfect example of why it is so difficult to make vast, overarching conclusions about my mtDNA without knowing all the colors of my autosomal rainbow…The same holds true for all the other Clovis Anzick matches—who also had non-Native American haplgroups like E, L, H, T, U, J and K,—who did not make her list either.