View Full Version : Does Ancestry test different SNPs outside the US?

10-07-2017, 10:18 PM
Background Summary:

WeGene was able to resolve Y haplogroup H1b2 (ISOGG 2016 tree) for my 2 brothers but only resolve H1 for me. I couldn’t understand why considering that we all tested with and uploaded Ancestry raw files.

Was it because we tested a couple months apart? Or could it be because they were done at different labs (mine in Dublin and my brothers in the US)?

Using basepair information that WeGene provided I searched for position 2697625 in our raw files. My brothers’ files had AA, Mine had 00. Ok, so that probably explains why.



Out of curiosity I started looking for other SNP differences (using UltraCompare); specifically cases where one of us had valid alleles and the other had zeros.

I noticed that my file had significantly more instances of "invalid" alleles (than my brothers’). See screenshot below for an example (my file is on the right side).


Question Summary:

Does anyone know if Ancestry’s US lab tests more / different SNPs than their labs outside the US?

And if so, could this be due to different technology and / or regulatory / government restrictions (regulations)?

10-08-2017, 01:58 PM

I have tested in Europe and I have the same question.
ISOGG states that AncestryDNA tests 637,639 SNPs, 1691 Y-chromosome SNPs, 28,892 X-chromosome SNPs and 195 in mtDNA.
When I run diagnostics on my AncestryDNA kit on GEDmatch I get 460431 SNPs used by GEDmatch. If I sum SNPs resolved by GEDmatch I get 455158 SNPs. In Genesis I can see that my MyHeritage and AncestryDNA-kits overlap on 404968 SNPs, but it is over 68200 SNPs in MyHeeritage kit (they match my mother).
Then I check my Genesis results with my AncestryDNA kit and what I see is that there are several types of matches: 646 000, 652 000, 657 000, 658 000 (all figures rounded) overlap in SNPs and other type like 404 000, 405 000, 403 000 (all figures rounded) overlap in SNPs.
Genesis is quite new so I cannot conclude on some difference in submission time. What I see in my raw results is V 1.0 chip (Europe, done in the middle of Augsut 2017).
It seems like there are two test versions or two chip types.
Could anyone confirm my observations?

On other side, the position you have marked in your attached pics is resolved in my results:
rs747730726 24 2697625 G G