Update: 8-Oct-2014
Uploaded a new kit F999919 processed from Clovis Anzick-1 BAM file. Refer: New Clovis Anzick-1 kit in GEDMatch: F999919
This blog is a follow-up of my previous post, Clovis-Anzick-1 ancient DNA have matches with living people! If you haven't read my earlier post, I suggest reading that first.
In the above diagram, 2 segments from the daughter which matches with Clovis-Anzick-1 sample also matches the phased paternal kit of the daugther (PF338998P1). This confirms that 2 matching segments from chromosome 1 and 6 from the daughter are indeed IBD (identity by descent). This does not mean the rest are IBS (Identity By State) or random matches. Instead the rest are compound segments caused by pedigree collapse through endogamous marriages.
One might ask the IBD is just for 1 generation and the segment for the father could very well be compound segment. Yes, it may be because the father's kit is not phased. However, we can still compare the matches with each other. If we take the matching segments from Clovis Anzick-1 and compare it with how those segments match each other, it is much evident that these segments are passed down to those matching people from a recent common ancestor.
![]()
This confirms that the matches are indeed valid and IBD.
Veronica (FN111284) her phased kit (PFN111284P1) matches perfectly at 5 cM/500 SNPs threshold with largest segment 8 cM.
![]()
This proves 3 segments matching the Clovis-Anzick-1 sample are IBD (Identity by Descent) through her father (paternal kit).
If your kit is phased, and you and the phased kit have a matching segment with Clovis-Anzick-1, please let me know and I will be happy to post it here which helps to confirm IBD segments.
As mentioned earlier, the fact that most of the Y-haplogroup for the first 15 matches to kitF999912 F999913 is Q1a3a* and the haplogroup of the Clovis-Anzick-1 is also Q-L54 (which is Q1a3a), confirms the paternal lineage.
If the matches are valid with IBD segments, how valid are the generations? Are they truly recent cousins? If we take only the IBD segments and based on IBD data of known relationships, the relation can only be very recent cousins.
If we know the matching segments are IBD and the relation is at recent cousin level, which could just be last century, how could it be 12500 years old and considered ancient? Well, that's an interesting question the authors of the paper needs to answer. Clearly something is wrong here. Irrespective of what went wrong, the conclusion is much evident.
Clearly, an IBD segment of 5 cM above 500 SNPs with total IBD segments around 10+ cM cannot be 12500 years old. This is a fact and can be verified using known relationships in families and DNA companies are using these benchmarks all along for showing genetic matches. This fact is more than enough to conclude that the Clovis-Anzick-1 sample is not actually ancient. My best guess is, the infant boy's sample is just from the last century and it was wrongly labeled as 12500 years old or the sample got contaminated.
This is purely my opinion on what I can see but I could very well be wrong.
Uploaded a new kit F999919 processed from Clovis Anzick-1 BAM file. Refer: New Clovis Anzick-1 kit in GEDMatch: F999919
This blog is a follow-up of my previous post, Clovis-Anzick-1 ancient DNA have matches with living people! If you haven't read my earlier post, I suggest reading that first.
Quick Recap
Just a quick recap, I processed the raw data for Clovis-Anzick-1 and uploaded into GEDMatch and to my surprise, there are matches as near as 3rd to 4th cousins. Now, that's a real problem because, the matches are to a DNA sample older than 12500 years. This is practically impossible and very mysterious. I will investigate step-by-step and see what are all the possibilities and failure points, which could solve the problem. But before that, we need to be absolutely sure that these matches are indeed valid. From the matches, I requested for phased kit and I indeed got one - Thanks to Mario Diaz and Veronica.Phased Matches
Both Mario Diaz (F349738) and his daughter (F338998) match Clovis-Anzick-1. But for his daughter, the threshold had to be little lowered 5 cM/500 SNPs. This is acceptable because his daughter is the next generation. The phased kits are PF338998M1, PF338998P1 - since phasing is done on the daughter's kit. |
| Phased segment matching Clovis-Anzick-1 |
One might ask the IBD is just for 1 generation and the segment for the father could very well be compound segment. Yes, it may be because the father's kit is not phased. However, we can still compare the matches with each other. If we take the matching segments from Clovis Anzick-1 and compare it with how those segments match each other, it is much evident that these segments are passed down to those matching people from a recent common ancestor.
This confirms that the matches are indeed valid and IBD.
Veronica (FN111284) her phased kit (PFN111284P1) matches perfectly at 5 cM/500 SNPs threshold with largest segment 8 cM.
This proves 3 segments matching the Clovis-Anzick-1 sample are IBD (Identity by Descent) through her father (paternal kit).
Triangulation
Some matches with Clovis-Anzick-1 can also be triangulated. |
| Triangulation with 3 people with each other and Clovis Anzick-1 at 7 cM/700+ threshold. |
 |
| Triangulation between Clovis Anzick-1, M193252 and M174237 |
 |
| Triangulation between Clovis Anzick-1, FN111284 and A778817 |
If your kit is phased, and you and the phased kit have a matching segment with Clovis-Anzick-1, please let me know and I will be happy to post it here which helps to confirm IBD segments.
As mentioned earlier, the fact that most of the Y-haplogroup for the first 15 matches to kit
If the matches are valid with IBD segments, how valid are the generations? Are they truly recent cousins? If we take only the IBD segments and based on IBD data of known relationships, the relation can only be very recent cousins.
If we know the matching segments are IBD and the relation is at recent cousin level, which could just be last century, how could it be 12500 years old and considered ancient? Well, that's an interesting question the authors of the paper needs to answer. Clearly something is wrong here. Irrespective of what went wrong, the conclusion is much evident.
Conclusion
The unexpected result causing this mystery matches could very well be due to some mistake or contamination of Clovis-Anzick-1 DNA sample itself. Roberta had blogged Analyzing the Native American Clovis Anzick Ancient Results where she shares her views on it. While I mostly agree regarding the matches having more Amerindian in admixture, the validity of the age as 12500 years old is never questioned. Even if we assume all the matches are due to endogamous marriages, 12500 years is impossible.Clearly, an IBD segment of 5 cM above 500 SNPs with total IBD segments around 10+ cM cannot be 12500 years old. This is a fact and can be verified using known relationships in families and DNA companies are using these benchmarks all along for showing genetic matches. This fact is more than enough to conclude that the Clovis-Anzick-1 sample is not actually ancient. My best guess is, the infant boy's sample is just from the last century and it was wrongly labeled as 12500 years old or the sample got contaminated.
This is purely my opinion on what I can see but I could very well be wrong.
