The newest one, K15 Eurogenes. Is the most accurate Admix run on Gedmatch. when it comes to the Oracle.
Printable View
The newest one, K15 Eurogenes. Is the most accurate Admix run on Gedmatch. when it comes to the Oracle.
Agreed. The other calculators place me weirdly, but K15 Eurogenes agrees almost completely with McDonald and my ancestral background.Quote:
Originally Posted by Graham
I'm going to have to go with the Eurogenes EUtest? Why? Because its oracle gave my mom 2.4 genetic distance to West & Central Germans
V3 and K12b are the best Dodecads, no? Anyway, I really love the new Eurogenes K15.Quote:
Originally Posted by Atlantic Islander
I think you should give identifying with the majority of your actual paper trail ancestry a try.Quote:
Originally Posted by Herr Hesselink
Not a fan of V3, I get France before Spain; twice.Quote:
Originally Posted by FaerieQueene
Yeah, the newest one is pretty good.Quote:
Originally Posted by Graham
I see. But don't you think the Admixture Proportions are pretty good?Quote:
Originally Posted by Atlantic Islander
Yeah, the numbers are okay, but oracle is a mess.Quote:
Originally Posted by FaerieQueene
None because there isn't a test that pinpoints specific genes from specific populations. If the tests cannot distinguish between Polish, German, or French then they aren't very accurate. Every single test gives me different results. Some say I'm mostly French, others mostly German, some Danish and Basque, some say there's Italian, other's say there's Cypriot...So the point is you cannot use the tests to determine Specific locations that your ancestors came from. Just because one test says on Oracle that you are 7% Serbian doesn't actually mean that 7% of your ancestors came from Serbia. So the tests are no good, they just tell me what I already know, that I'm mostly Northern Euro and Part Southern Euro.
Admixture wise Dodecad/Eurogenes K12b and HarappaWorld.. Oracle wise Eurogenes K15 and HarappaWorld..
Well you have to apply a level of interpretation. If you look at it logically, it generally makes sense unless there is a problem with the calculator itself.Quote:
Originally Posted by JQP4545
They can give you an idea of, well lets say you are a quarter Dutch, a quarter German, a quarter Italian, and a quarter Polish. The tests will always say that you are about 25% Southern and 75% Northern European, but it will always give you some random populations. Maybe it will say you are 50% French, etc... The tests don't pinpoint specific nationalities. So how are they anymore useful than the ones you pay for on 23andMe which give you specific nationalites? Some test might say you are 2.33 percent Mozabite when really you are Italian, so what is the point?
Well if you have a more complex ancestry the answer is going to be more complex too, but the answers do make logical sense. The populations it gives are based on their average scores compared to yours. So while it may for example say someone who is for example 75% British 25% Italian is actually French, but this makes logical sense.Quote:
Originally Posted by JQP4545
Don't the French have unique genes? I don't think they are 75% British and 25% Italian. Certain ethnicities have unique diseases and physical features, so a good test should be able to distinguish between these groups.Quote:
Originally Posted by Jackson
Well you have to understand that everyone is on a genetic spectrum. These programs look at genetic components, and each group has a particular average composition, so if you are very close to their average composition you will show up as being similar to them, whether that's because that person is actually French, or because they are a mix of people that averages out close to the French average, it is not always able to tell. Although usually a mixed person will find that they are represented as (for example) something close to 75% British 25% Italian on the mixed mode, rather than just French. I would recommend the most recent Eurogenes tests as they have the best track record, especially when it comes to the mixed-mode oracle.Quote:
Originally Posted by JQP4545
Basically your genetic profile is a numerical problem that the program solves by comparing it to other identities (country averages in this case) in it's database. It's purely a mathematical thing, and looks at ancestral components rather than diseases and physical features - which are usually less useful in distinguishing groups of people. I know that a good genetic calculator can place a person in their country without much trouble, and often the mixed mode is very close to reality.
Yes the french are their own ethnicity, however you being 75% brit and 25% italian the closest population to your average are the french that's why it shows up. Genetically it also makes sense as french ancestry would be like british but with a pull towards southern-Europe.Quote:
Originally Posted by JQP4545
The Eurogenes tests have been much more accurate for me than 23andme.Quote:
Originally Posted by JQP4545
I think Harappa world is the worst for people of mixed southern European and Northern European ancestry..
Look at my Harappa results in mixed mode:
Using 2 populations approximation:
1 50% ashkenazy-jew +50% utahn-white @ 3.865
2 50% ashkenazi +50% french @ 4.000
3 50% ashkenazy-jew +50% british @ 4.071
4 50% italian +50% romanian-a @ 4.134
5 50% italian +50% serbian @ 4.304
6 50% ashkenazi +50% utahn-white @ 4.436
7 50% ashkenazy-jew +50% utahn-white @ 4.565
8 50% ashkenazi +50% british @ 4.712
9 50% ashkenazy-jew +50% n-european @ 4.923
10 50% hungarian +50% tuscan @ 4.996
71253 iterations.
Quote:
Originally Posted by The Exiled King
From Eurogenes I like Eurogenes EUtest k12b and k13 the most.
From dodecad I like V3, world 9 and k12b the most.
MDLP k12 and k10 is good enough.
Dodecad k12b.
Eurogenes K13 is quite good, but EUtest version 1 is probably the most accurate, from what i can tell.
EUtest V1 imho, since it's not an option i voted for MDLP, which is also relatively accurate.
Harappa DNA, It actually has a seperate Onge component
What's the best calculator to take for a European - the most accurate?
I'd recommend Eurogenes, but for two reasons in particular:Quote:
Originally Posted by Wulf Talented
1) The Oracle doesn't suffer from the 'calculator effect' as most others. This basically means that people who weren't in the original run get pulled towards the middle of the populations, that caused a lot of confusion for a lot of people a little while back. In this case only the components are reliable, but you have to compare them with other people of known ancestry who were also not in the run - but in that respect it can be good. For example i really like some of the MDLP calculators and Dodecad ones, but they are only useful in this regard, not with the oracle.
2) At the higher Ks it breaks up some of the more closer-knit components like in the NW, East, SW of Europe for example.
The Eurogenes EUtest is probably the most accurate for all members of my family, but it's worth trying them all to see what they say (although bear in mind the calculator effect for other projects, which seems to be real from my experiences).
That said, it's worth reading the blogs of Dienekes, MDLP, Eurogenes and Harappa. Dienekes does a broader anthropology blog which relays really useful information and his take on them, and so does Eurogenes blog, but both have genetic projects (Dienekes' is called Dodecad).
Thank you very much for the response Sir,Quote:
Originally Posted by Jackson
I have just re-taken the Jtest and EUtest. I took them once before and posted the results on here somewhere, but with a bit more understanding about population averages and etc, It seems like I might suffer the calculator affect... I come out slightly higher as North Central European for the J and EUTESTS than what my ( belonging to ) population average shows...
EUTEST ( original ) - NORTH-CENTRAL_EURO 34.20%
Population average for English - 29.67%
EUTEST v2 K15 NORTH-SEA 44.37%
Population Average for Southeast English - 35.52
JTEST NORTH-CENTRAL_EURO 34.01%
Population average for English - 29.44
I don't think you suffer the calculator effect for EUtest/Jtest because it doesn't occur on those tests, but i noticed before you are very high in this one component. Sometimes that happens i guess. It's probably going to make you look more Scandinavian (rather than Polish) than you'd expect but it's something youll have to bear in mind. Just think about it as - There's a lot of variation around the population averages, you just happen to be a small percentile that get significantly above average for this, although it should balance out overall in that you get lower than average for some other components i guess. For example both my grandparents have about average for that component, my father has about 2% above average, and the phased data file has about 3% above average (in the EUtest for example). It's interesting that the North Sea component in EUtest V2 K15 has a peak in Orkney, so it may just be saying that you have something that makes you look genetically more like the average Orcadian than most other English people do. Graham (a Scottish member here) also has a very high amount of that component on EUtest V2 K15, and i think his ancestry is from mainland Scotland, northern England and Shetland for the most part.Quote:
Originally Posted by Wulf Talented
What you'd expect from the calculator effect i think is that you would be pulled towards the centre, so a lot of people from the British Isles come out something like 70% British 30% Hungarian, even if they are typical in other tests not affected.
It's very confusing, I understand that my Polish ancestry will pull me closer to continental Germanic populations, my Oracle results always have me Germanic predominantly. I once asked Davidski about it, my population distances were really distant in terms of genetic similarity - my first population being #1 Orcadian @ 8.708, so it's due to the Polish ancestry muffling it all up.Quote:
Originally Posted by Jackson
It's true that where I am significantly higher in North Euro results, my Atlantic/med results drop below the average.
An example using the EU vs K15 Test again -
Population average ( South east English ) - 29.86
Atlantic 20.62%
So does that mean the 34% North Euro is not legitimate... or it is but only due to Polish ancestry pulling it towards Poland?
It's been really difficult to distinguish my genetic make up so far due to this Polish ancestor. I'm not sure what the Admix calculators really do, are they suppose to show your genetic make up from specific areas in the world, be it distant or ancient ( I know that the bloggers try to get it through to the people that the populations shouldn't be taken seriously )? I'm trying my best to determine The Polish genetic make up, but 23andMe labelled him nonspecific Euro at 10% ( I assume it's him, rest is North Euro and like >0.1 North African, 0.1% unassgined, ), and FTDNA Family finder just isn't that good IMO. These calculators haven't helped, haha!
I think they are probably not helpful because your results don't seem what you'd expect for your ancestry. In theory your Baltic and Eastern Euro should be higher than average, your Atlantic and Mediterranean lower than average, and your North Sea/Central Euro about the same as average or lower. I'd have thought it's possible that your English ancestry is unusually northern/northwestern, and therefore an eastern pull is harder to see - but in reality the calculator still picks up an eastern pull, even if it just uses 'Orcadian' as the base population a lot of the time. But yeah it could be that both your English and Polish ancestors weren't typical of the average, so the result (British Isles and a pull to the east) can still be seen, but it's expressed in a different way perhaps. Either that or you just happened to get unusual results, would be useful if you had other family members go through it to see if it's easier to see in them other than yourself, because of course two people with the same ancestry (unless they are genetically identical twins i assume) can come out a bit differently on these calculators despite having the same ancestry, two different results of mixing the same things i guess - maybe you just got an unusual result rather than a more expected one?Quote:
Originally Posted by Wulf Talented
It basically seems to pick up that you are atypical, and overall have a pull to the east, although because the distribution of your components is not what you'd expect of a mix of someone close to the English population average and the Polish population average, it has trouble identifying these two populations as the actual mixture.Quote:
Originally Posted by Jackson
I think in that case it may be best to have my father and mother tested, and then throw their results into the the Admix calculators. The Polish ancestry lies on my mother's side and my father I suspect to be entirely English for his ancestry lies soloely around South-East England. ( at least as far back as the late 1700's via a family tree.
Thank you Jackson!
Wulf Talented you score high North Sea in K15. Where in England are you from? Because it scores higher in the more Norse influenced Brits.
My English ancestry for my father lies as far as I can trace back to South to South-East England. his ancestors were dotted between and around Kent, Sussex and the Isle of Wight. My Mother's English ancestry aside her Polish ancestry is Grantham, Nottingham, York, Northampton and Cambridge.Quote:
Originally Posted by Graham
I was born in Kent, then ironically moved to Lincolnshire 5 years ago!
MDLP because it's the only one that gets it right for me.
It's not on the list but K36 from Eurogenes is probably my favourite.
Sure!Quote:
Originally Posted by safinator
Eurogenes K13 and Dodecad V3 :)
K13
How come?Quote:
Originally Posted by HellLander87
How come?Quote:
Originally Posted by callmeswarthy