Thursday, October 2, 2014

Mezmaiskaya neanderthal kit from GEDMatch removed

I am revisiting some of the kits uploaded to GEDMatch with less SNPs common with DNA testing companies. I found one such kit which is Mezmaiskaya Neanderthal uploaded as kit F999909.

You can see how much less the number of SNPs are present using GEDmatch DNA file diagnostic utility. Since it doesn't add any value, I am removing it from GEDMatch. But the processed files are still available to download from

Mezmaiskaya Neanderthal Kit 
This is to improve the quality of Ancient DNA kits uploaded to GEDMatch.

Monday, September 29, 2014

The female Y-chromosome of Neanderthals and Denisovans.

When Neanderthal Genomes was released it had a Y-Chromosome. No one knows why but many, including the authors had suggested it could be contamination by a research worker. However, such suggestions will lead to rejecting every experimental results if the outcome is opposite of what we know and expect. I do not want to reject the Y-Chromosome as contamination without proper investigation. Not only Altai Neanderthal, but also both Mezmaiskaya Neanderthal and Denisova also have Y-Chromosome inspite of them being females.


Altai Neandertal inhabited the same Denisova cave in Altai Mountains where the Denisovan fingerbone was found. The Y-DNA for Altai Neanderthal is R-L51 and Denisova is R-L23. R-L23 (R1b1a2a) is the parent haplogroup of R-L51 (R1b1a2a1).
Mezmaiskaya Neanderthal however, is found nearly 5000 kms from Denisova Cave and has no matching Y-DNA. However, it does have some SNPs mostly from Chromo2 set, that are not yet in the tree. When I compared these SNPs with my own DNA, I match more than 60% and less than 10% I am negative, which suggests the missing y-SNPs may yet to be in the Y-Tree.

Autosomal Comparison

When I compared Altai Neandertal with Denisova at 2 cM/500 SNPs with mismatch-bunching limit at 500 SNPs in GEDMatch, I do have one significant match. However none between those two and Mezmaiskaya Neanderthal.

Altai Neanderthal and Denisova

So, I decided to investigate further esp., on this particular segment and see how many no-calls were there and confirm if the segment is trust worthy. What is uploaded in GEDMatch is a filtered version of the actual file because, DNA testing companies like FTDNA, 23andMe and Ancestry don't test all SNPs found in these ancient samples. However, for testing this segment, I am going to use the complete SNP set.

There are a total of 26150 SNPs in the complete set for this particular segment. This is more than enough to confirm if they are related through this particular matching segment. When I did a initial comparison, there were 150 differences which includes no calls and errors. Because there are too many SNPs in comparision and there were only 150 that differs, we can confidently ignore the no-calls. This leaves with 120 errors. It is important to note that there will also be errors in testing which must also be considered. Considering the fact we are comparing 26150 SNPs and the error is only 0.45%, this segment is definitely from a common ancestor and they are related for sure. The odds of having such a match accidentally is next to impossibility,

The data used for analysis can be found here.


Based on the fact that Altai Neanderthal and Denisova are related autosomally and their both Y-DNA lineage are also same, I can say for sure they are indeed closely related and requires further investigation. I also believe Neanderthals and Denisovan females did had Y-Chromosome.  Are they really females or something that we never knew of ?

Sunday, September 28, 2014

Mal'ta MA-1 ancient DNA have matches with a few living people on X-Chromosome!

I posted the processed sequence of an ancient genome of individual (MA-1), from Mal’ta in south-central Siberia in formats familiar to genetic genealogists which I made available here and uploaded to GEDMatch as kit# F999914. I also posted a blog analyzing what i found initially. I wasn't expecting any surprise matches but I was wrong again. Unlike the Clovis-Anzick-1 sample having matches on autosomal DNA, the Mal'ta MA-1 sample is having matches on X chromosome. My immediate thought was wow! These matches, although small deserves further investigation.


GEDMatch 1-to-many matches for Mal'ta Sample
So, I decided to Triangulate. Wonder what? I was able to successfully triangulate at 5 cM/500 SNPs (considering the fact the sample is ancient).


Triangulation between M175228,M133519 with MA-1 sample

Triangulation between M175228, M184156 with MA-1 sample

Triangulation between M061014, M000556 with MA-1 sample

Let me know your thoughts and comments!

Analyzing La Braña-Arintero Ancient DNA

Approximately 7,000-year-old Mesolithic skeleton discovered at the La Braña-Arintero site in León, Spain, had been sequenced to retrieve a complete pre-agricultural European human genome. I converted the raw sequence reads supplied in the scientific paper to formats familiar with genetic genealogists and uploaded here and also to GEDMatch as kit# F999915


Y-Haplogroup: C-V183


  • DYS638 = 11
  • DYS461 = 12

Based on my experience, the values could be offset by 1.


The mt-DNA Haplogroup for the kit is U5b2c1


The kit has a telomere length of 5.83782 which may suggest the boy's age was roughly around 10 to 12 years at the time of his death.

Plotting telomere length of 5.878 kb
(Image adapted from


Based on MLDP-K23b calculator, below is the admixture.

MLDP-K23b admixture for La Braña-Arintero Ancient DNA

Runs of Homozygosity

RoH reveals the parents of La Braña-Arintero are not related to each other in his genealogical time frame.

I did one-to-one with my DNA and nothing in common. May be you have something in common. Let me know what you think.

Saturday, September 27, 2014

Uploaded New GEDMatch Kit for Clovis-Anzick-1

Clovis-Anzick DNA was uploaded to GEDMatch with FTDNA SNPs alone as kit# F999912, since I was unable to upload the complete file which is larger size and I didn't expect anyone to match because it is an ancient DNA. Interestingly, it had significant matches with living people which was unexpected. I did my own analysis of the matches which has triangulated and phased matches and I decided to investigate more. So, I decided to extract SNPs used by all DNA Companies (FTDNA, 23andMe v3 and v4, Ancestry). I created a new kit F999913 which has significantly more SNPs which replaces F999912.

Kits F999912 vs F999913

Looking at the GEDmatch DNA file diagnostic utility for F999912, Chromosome 21 and X are still below the GEDMatch threshold.
GEDMatch DNA Diagnostic Utility for F999912

However, the new Kit F999913, all SNPs are above the GEDMatch threshold.

GEDMatch DNA Diagnostic Utility for F999913

Once the batch processing for F999913 is complete, I will remove the kit F999912.  I had removed the kit F999912 and one-to-many comparison for F999913 is available in GEDMatch.