Printable View
I recently did an autosomal test with National Geographic and I was wondering how accurate the results that come back are? The smaller percentages in the result do they fall within margin of error? I'm just curious, I'm not 100% sure about the authenticity of autosomal testing. Sorry if I'm not making sense, hopefully someone can shed some light on whether to trust these results or not.
Thanks
To answer your question, I assume that you are wondering about some minor ancestry percentages in your test results. There are a number of factors to consider but I'm not going to bore you with statistics. What is important to consider is the direction of gene flow. For example, if an individual's result states 3% East Asian, this could be due to a number of scenarios:Quote:
Originally Posted by Barantes
1. There is 3% East Asian ancestry.
2. There is European ancestry that is also found in East Asia at higher frequencies, programmes will take the higher frequency to mean that this ancestry originated in East Asia, although we cannot say this for certain.
3. There is 3% ancestry from an unknown source that has managed to spread its genes to East Asia where it's frequency is sufficient to suggest East Asian ancestry, however this would be inaccurate.
Good luck with your analysis!
That was really helpful actually thank you. I read that there is also a possible margin of error I think it was called "static noise" or something I'm not too sure what they referred to it as.
They're extremely accurate if done properly and interpreted correctly.
the best I have found are Eurogenes one...JTEST and EUTEST
the newish Eurogenes K36 has many issues unanswered and seems incomplete ........ie..missing certain ethnic people
I don't think that they are 100 percent accurate, but they can tell you some kind of picture about yourself. Three of my grandparents are Slovak and my fourth is a mix of the british isles and central European countries. I took an autosomal test with ancestry.com and I came back as 29% Scandinavian, 63% Eastern European and 8 percent uncertain. I had my parents take the same test and my mother who is half Slovak and half(the fourth grandparent) came out as 54% Eastern European and 43% central European. My Father who is full Slovak came out as 45% Eastern European, 34% Central European and 21% Finnish Volga Ural. As you can see there is some confusion. How did I get so much Scandinavian when My parents have none. Well if we think about it a little maybe its not so confusing. We know central Europe (Germany and Franice in this test)is a mix of Germanic people from Scandinavia who came south and mixed with celtic type peoples. So Autosomally my parents showed up as having Central European because they have that pattern of a mix of celtic and Germanic genes. However maybe I only got the Germanic segments of those genes which best matches me with Scandinavian people. Also the Finnish/Volga Ural, a region which stretches from finland to the Caucasus has a lot of Slavic genes wich probably my father has in common with eastern Slavic people, and that is probably the 8 percent% unkown in me. This is probably the best way to look at autosomal testing. It is somewhat accurate but not perfect. It depends heavily on the testing companies database, and what the conclude as a homogenous region.
Thats not strange becuse slovaks is a mix of slavic and central european (celtic&germanic peoples)Quote:
Originally Posted by silkyslovanbojkovsky
Yes I never said it was strange that slovaks have central european dna. I was just trying to explain how I ended up as showing 29% scandinavian dna although my parents didnt show up as having any.Quote:
Originally Posted by Toscano
There is a new product to predict your ancient commencement in Europe or elsewhere
Go to Gedmatch and use eurogenes K9b to determine your data, then place results below and use #9 for reference numbers
http://mendel.fat.glam.ac.uk/project...2/personal.php
you will get a match like mine below ( initially mine had central france but it was a beta test)
http://img405.imageshack.us/img405/3682/lcny.png
Uploaded with ImageShack.us
below are my results from K9b - first time over 5% for SW-Asian
Population Southwest_Asian 12.09% Native_American 0.54% Northeast_Asian - Mediterranean 29.67% North_European 55.77% Southeast_Asian 1.42% Oceanian 0.28% South_African - Sub-Saharan_African 0.25%
23andme is a great adventury. Family Finder is a silly joke. The say Bulgarians are like 70% or 80 % British or French and 20% or 30 % Egyptian or Druzeh he he he he
In this GPS system i notedQuote:
Originally Posted by Yaan
The K9b eurogenes uses the system that Nat Geno 2.0, so South-west Asian incorporates -
Arabian peninsula
Irak
Iran
Pakistan
India
so the range is huge
I would agree with the rest and say that autosomal tests are pretty accurate, as long as they are run with the admixture calculators found at GEDmatch. Don't forget that they are testing a huge number (hundreds of thousands) SNPs and they use a large number of reference populations around the world. The only limitation I see is that some times these reference populations are too small and they do not actually represent the population that they are supposed to do. Now for the default autosomal calculators of FTDNA, 23andme and Geno 2.0, all these are indeed problematic (the 23andme one probably less than the others), especially if your ancestry comes from southeastern Europe. For people with western/northern Europeans (including Americans/Canadians/Australians), these default calculators should work fine however, but they will not give you the detail that the GEDmatch calculators will.
interesting article on Audna - noise threshold
http://www.y-str.org/noise-threshold-on-atdna-matches/
I just got paper records bringing my maternal line ( for DNA purposes) one further generation back with the name Zotton
The name in different Italians sites for ancient surnames states either - Lombard
Famiglia friulana di antica e nobile tradizione propagatasi, nel corso dei secoli, in diverse regioni d'Italia. La cognominizzazione รจ anteriore al XV secolo e dovrebbe derivare dal nome longobardo Zotton. Ne abbiamo alcuni esempi nel Codice Diplomatico Longobardo: nell'anno 715.
or Carnico , which means from the carni tribe -----The carni tribe from my information was a Gallic tribe from Gascony area which invaded north-east italy around 400 to 500 BC .
The one thing I want to discuss with this is ------I am prominent in any gedmatch Admixture sites which have french basque
or
1
North_Italian 49.82% 2 French_Basque 8.88% 3 Spanish_Andalucia 8.42% 4 Chuvash 7.55% 5 Adygei 6.98%
to place a few examples
1
Austrian
37.78% 2 French_Basque 26.90% 3 North_Italian 20.53% 4 Belorussian 4.59%
Zotton is clearly an extension of original Zotto in which you can get zottin, zottolo, del zotto etc
conclusion- maybe these gedmatch admixtures have more info that what I originally thought
What do you mean by correctly exactly? Are the models built for certain groups of people?Quote:
Originally Posted by polako