PDA

View Full Version : Gencove



Carl Graham
15-09-17, 06:32
Has anybody else tried Gencove yet ? To me it seems consistent with most other reports from most other companies.Maybe a tiny bit less southern European and a tiny bit more Scandinavian would be more in line with the average result , but I really can't quibble.

9268

Dibran
15-09-17, 20:47
Has anybody else tried Gencove yet ? To me it seems consistent with most other reports from most other companies.Maybe a tiny bit less southern European and a tiny bit more Scandinavian would be more in line with the average result , but I really can't quibble.

9268

Is it a 3rd party tool or a full testing company? Is this all they provide?

noman
16-09-17, 03:11
http://i.imgur.com/aYPzziC.jpg

Dibran
16-09-17, 10:34
Eastern Mediterranean
Reference populations: Malta, Sicily, Cyprus, Greece, Albania, and Bulgaria.


Southwestern Europe
Reference populations: Southern French, Spanish, Basque, and Sardinian.




My Results:


23andme:


https://s26.postimg.org/hxn0n4czd/gen23.jpg


Ancestry:


https://s26.postimg.org/wuvho4q7t/genan.jpg


My Father:


https://s26.postimg.org/pg65vr4c9/genr.jpg

kingjohn
17-09-17, 11:57
my results { close to dna land not so much like my origins and my heritage}
that detacted 9-11% slavic genes and 0% here ....
9277

my results
33% aschenazi
23% south west Europe
12% east med
17% middle eastern
8% Anatolia Caucasus Iran
8% central India

Promenade
17-09-17, 12:48
{ close to dna land not so much like my origins and my heritage}

The CEO of Gencove is Joe Pickrell, part of the DNALand team

Stuvanè
17-09-17, 15:44
Done. Even though some of my friends have detected different values ​​after having uploaded their raw dates several times (me too)https://uploads.tapatalk-cdn.com/20170917/9e7838c2d0c60ffa7cbea5e469238c8b.jpg

Sent from my SM-A300FU using Eupedia Forum mobile app (http://r.tapatalk.com/byo?rid=89698)

AdeoF
17-09-17, 18:22
I done it as well, but the info I got. But it does look a bit off however, but that could be me

9280

Angela
17-09-17, 19:58
I knew I shouldn't have used a calculator based on modern samples again. I hate to say it, because Joe Pickrell is a reputable person, but if someone was adopted and wanted to know their ethnicity, they certainly wouldn't learn it from here. Sorry.

Southwestern European 74%
Eastern Med (Sicily, Malta, Greece, Albania, Bulgaria) 8%
Caucasus 8%
Ashkenazi 2%

Again, I have tested with 23andme and I don't have one single instance of IBD sharing with Ashkenazim or Jews of any variety, not even those who are half, last I checked, unless the other half is Italian. So, wrong.

I only share 8% with Sicilians, Greeks, Albanians, Bulgarians? If there are no Italian reference samples, as in some of Kurd's runs, I come out as either Albanian or Bulgarian. Also, on some Oracle runs, the Greeks and Albanians show up before the Sicilians, although my first scorers are always the northern Italian and Tuscan ones. The Piemontese/Emilian/Ligurian one always comes in first if that's included. Plus, why would someone have a cluster where Albania and Bulgaria are called Eastern Med? Are they picking up what Kurd called, in the ancient samples, the eastern farmers? However, I get 28% of that, not 8%.

The 8% Caucasus might make sense as "additional" Caucasus if they're placing the centrum for both central (i.e. Italians) and southwestern Southern Europeans in southern France and Iberia. According to the comparison with ancient samples in geneplaza I have rather high steppe for an Italian, probably higher than Iberians, and definitely higher than the southern French sample.

I mean, I know I've said there is probably a lot of overlap between the southern French and northwestern Italians, and I certainly have gotten a lot of matches on Oracle with Spaniards, but 74%? I don't know, maybe, I guess.

I think I'll stick with comparisons to ancient samples.

Ed. Also 8% northern and central Europe.

kingjohn
18-09-17, 08:38
angela you score high south west Europe here
because even though it is the same algorithm as DNA land
here the is no south /central european cluster like in dna land{ reference bergamo Tuscany}
so many of your north central Italian genes go to the south west european category
in my opinion it is just a bad version of DNA land same algorithm just less references
and very different from my origins 2.0 and my heritage algorithm
kind regards
adam

Northener
18-09-17, 09:14
angela you score high south west Europe here
because even though it is the same algorithm as DNA land
here the is no south /central european cluster like in dna land{ reference bergamo Tuscany}
so many of your north central Italian genes go to the south west european category
in my opinion it is just a bad version of DNA land same algorithm just less references
and very different from my origins 2.0 and my heritage algorithm
kind regards
adam


Yes indeed king John! Theres is also not a kind of northwest cluster. DNA Land has got a Northwest cluster, my score on that cluster 91%. But on gencove, because it has no NW cluster it spreads some more:

https://www.mupload.nl/img/vgrrnn7tfofhh.png

Maybe they have to finetune more?

davef
18-09-17, 12:57
What a stupid and pointless test. It's like a bad version of calculators that have already been around for years now; its like a failed attempt to reinvent the wheel.

Never heard of joe pacarel

davef
21-09-17, 18:45
Angela's scores add up to 92 percent....

Angela
21-09-17, 19:01
Davef is correct. The box doesn't scroll down correctly. I had to click on the geographic circles. I have 8% northern and central Europe. I'll correct the post.

davef
21-09-17, 19:14
Makes sense! ^^

Eastern Mediterranean
Reference populations: Malta, Sicily, Cyprus, Greece, Albania, and Bulgaria.

I have no idea why these are in the same boat.

Angela
21-09-17, 19:45
Makes sense! ^^

Eastern Mediterranean
Reference populations: Malta, Sicily, Cyprus, Greece, Albania, and Bulgaria.

I have no idea why these are in the same boat.

Now that I've thought about it more, it's probably shared more heavily "eastern" farmer ancestry versus "western" farmer (actually, probably Iberian Chalcolithic) ancestry). That's why I only get 8% "Eastern Med" (to Northener's 6%, for example), compared to my much, much higher "Western Med" score.

It's certainly has nothing to do with "Middle Eastern", or "Levantine", or "Arabic",which has it's own cluster, as does Caucasus like.

davef
21-09-17, 21:21
I know, it covers south Italy, balkans, Sicily, the Aegean, and parts of turkey.
But why are you getting 74 percent southwestern here and only 39 percent western farmer from Kurd's calculator? That's a huge discrepancy.

Angela
21-09-17, 22:29
I know, it covers south Italy, balkans, Sicily, the Aegean, and parts of turkey.
But why are you getting 74 percent southwestern here and only 39 percent western farmer from Kurd's calculator? That's a huge discrepancy.

You can't compare these two tests exactly one to one. These are modern clusters, and Kurd's are ancient samples. All of these modern clusters are going to have Yamnaya (with CHG/Iran Neo in it), plus any additional WHG, as well as early Neolithic farmer. All modern Europeans are a mix of those groups. It's just that the proportions are going to be different. The "Eastern Med" just probably contains a lot more eastern farmer, less Yamnaya, and less WHG as we know from other analyses is the case.

As was pointed out above, if there were a south/central cluster here (based on Northern Italy/Tuscany) as there is on some other tests, I wouldn't get so much southwestern.

This isn't wrong, it's just a different way of looking at things. However, I don't think it's very useful for newbies who don't know their ancestry specifically, or who aren't quite sophisticated in their understanding of population genetics. They're just going to get confused.

Tomenable
21-09-17, 23:26
Here are my ethnic Polish results in Gencove, DNA.Land, GenePlaza and DNA Tribes:

DNA.Land:

55% North Slavic
36% NW Euro
8.2% Balkan
1.3% SC Euro

GenePlaza:

58% North Slavic
29% NW Euro
6.2% SW Euro
5% Mediterranean
1.4% Ambiguous

Gencove:

54% NE Euro
39% N/C Euro
5% East Med
4% SW Euro

DNA.Tribes:

48.2% Belarus
35.8% Denmark
9.8% Slovenia
5.5% Sweden
0.6% Lithuania
0.1% Other

And I can tell you that these algorithms are probably biased, because it seems that they overestimate North-West European admixtures and underestimate North-East European admixtures. Probably they count too much of Central European genetic variation as part of NW cluster, and not enough of it as part of NE cluster. They do this so that their British or Irish customers can get ~100% North-West, and avoid getting North-East.

One example of a British-American result - 98% NW Euro:

http://img.photobucket.com/albums/v336/naudiz/genocove_zps5159wqvf.png

Albanians tend to get 100% Eastern Mediterranean and no any Slavic.

And where is my 5% from? Ancient Albanian migrations to Poland?

Tomenable
21-09-17, 23:27
My results:

https://i.imgur.com/Mc8S0T3.png

davef
22-09-17, 07:51
My results:
https://i.imgur.com/Mc8S0T3.png
It's as meaningless as the test itself. Since Angela is 39 percent western Neolithic, what is the rest of her "southwestern European" like? We'd never know without the new Kurd calculator, right? These categories are too vague and broad.
I know, I'll create a davef calculator which has the following categories:
Northern European (literally, the northern half of Europe)
Southern European Greco-Romano-Iberia
Middle East and Caucasian (spreads all across turkey, Arabia, levant and Iraq/Iran)
North African (morocco to Egypt)
Sub Saharan (Nigeria southwards)
India/east Asia and Australian
Let's be even more vague:
Human
Neanderthal

Why is this calculator even around? It's meaningless and uninformative...I wonder if it's a college project assigned to CS grad students...

Pax Augusta
22-09-17, 14:14
Why is this calculator even around? It's meaningless and uninformative...I wonder if it's a college project assigned to CS grad students...

I suppose because in this way they want to advertise their paid services. Gencove is a start up. It's just marketing.

New Englander
07-11-17, 20:08
Results.

Not sure why I get India, but I always seem to get a tiny bit. It may be from the Jews. Also, lots of SW Euro, taking from NW Euro. Might mean my Scottish ancestors mixed with the French in New Brunswick.

noman
07-11-17, 22:43
Looks like, there is a need for a new calculator.
I am eagerly waiting for Genplaza modern calculator. Kurd is working on it.

New Englander
09-11-17, 23:57
Really though, Im not sure why my South Western European is so high, its a 2:1 ratio compared to North Western European, that just does not coincide with the surnames from the family tree.

Gallagher
MacDougall
Prince (french?)
Coates (french?)
Nash
Bennett (french?)
Currier (French? )

Angela
10-11-17, 00:38
Why would you expect more northwestern if you're only 1/4 British? Even if you're half British, they're more than half "southern" as it is.

New Englander
10-11-17, 04:46
Northern and Central Europe

Reference populations: Norwegian, Orcadian, Scottish, British, Icelandic.

SO why would they be mostly South Western?

Im guessing a breakdown of the following:

Italian:
Eastern Med 37%
Middle East 4%
Iran 7%
South Western 2%

Jewish:
Ashkinazi 22%
Indian 3%

Canadian:
South Western 17%
North Western 8%

mwauthy
10-11-17, 20:28
Really though, Im not sure why my South Western European is so high, its a 2:1 ratio compared to North Western European, that just does not coincide with the surnames from the family tree.

Gallagher
MacDougall
Prince (french?)
Coates (french?)
Nash
Bennett (french?)
Currier (French? )

None of the reference samples for northwestern are from mainland Europe. They are all from UK and Scandinavia. So if you have ancestors from France, Low Countries, Germany, And Switzerland a decent percentage will be assigned to Southwestern because its the closest region. Ethnicity estimates are all about how you compare to the reference samples being used and have very little to do with where your ancestors were actually born.

New Englander
11-11-17, 21:55
I know, that is what I'm implying. Obviously a majority of the ancestry is NOT from Scotland, the UK, or Ireland.

Now, would it be possible for Native American genes to be identified as South Western Europe? Lack of samples from Mik'maqu populations may be a factor. Since French-Canadians have a high mixture of Native Ancestry in some areas, the lack of samples may lead to the assumption of a founder effect and labeled French? (South Western). Basically could say 4-5% of that 19% be misidentified Native American?

mwauthy
12-11-17, 01:54
I know, that is what I'm implying. Obviously a majority of the ancestry is NOT from Scotland, the UK, or Ireland.

Now, would it be possible for Native American genes to be identified as South Western Europe? Lack of samples from Mik'maqu populations may be a factor. Since French-Canadians have a high mixture of Native Ancestry in some areas, the lack of samples may lead to the assumption of a founder effect and labeled French? (South Western). Basically could say 4-5% of that 19% be misidentified Native American?

I’m 50% French Canadian back to 1650 and I don’t register any Native American on my various DNA tests. If you are part Metis then you will probably register some. However, the French Canadians in Quebec tended to marry other French Canadians.

I’m pretty sure that there is no correlation between Southwestern European and Native American. I get 27% Southwestern European on Gencove. This is due to there being no references samples for most of France. Many French Canadians came from La Rochelle. On a map you will see that that port is closer to Atlantic Spain than to Scandinavia or the Orkney Islands. So a lot of these Western and Central French genes get assigned to the Iberian Peninsula

Moi-même
12-11-17, 04:49
I know, that is what I'm implying. Obviously a majority of the ancestry is NOT from Scotland, the UK, or Ireland.

Now, would it be possible for Native American genes to be identified as South Western Europe? Lack of samples from Mik'maqu populations may be a factor. Since French-Canadians have a high mixture of Native Ancestry in some areas, the lack of samples may lead to the assumption of a founder effect and labeled French? (South Western). Basically could say 4-5% of that 19% be misidentified Native American?

Comparing my parents' and my sample (all of us should have North East Amerindians) with one of my grand-mother's sample (who should have only Europeans), the difference is trace amount of Asian DNA. Finnish, Siberian, Central Asian, South Asian, South East Asian, Oceanian, East Asian, a bit from one or another, depending of the test. Your unexplained South Asian, could be coming from Acadia.

New Englander
12-11-17, 05:54
Id like to see some North Eastern Native Americans results. I think there were two waves of natives into the Americas. The pre-Clovis and the Arctic Hunters. I'm wondering if they were a mixed population being so close to the Arctic, or if they were identical to the other Eastern Woodland tribes.

alexfritz
14-12-17, 03:29
47% Southwestern Europe
40% Northern and Central Europe
5% Eastern Mediterranean
5% Northeast Europe
3% Iran Caucasus Anatolia

https://i.imgur.com/DKBO8EO.png

davef
14-12-17, 03:34
47% Southwestern Europe
40% Northern and Central Europe
5% Eastern Mediterranean
5% Northeast Europe
3% Iran Caucasus Anatolia

https://i.imgur.com/DKBO8EO.png


Your other half is all that southwestern European that Angela got so yeah youre definitely 1/2 north Italian.

Angela
14-12-17, 04:01
Your other half is all that southwestern European that Angela got so yeah youre definitely 1/2 north Italian.

Absolutely not. I'll repeat. No one who is half Lombard and half German is going to get these scores. They'd be more "northern", i.e. higher northern Europe.

mwauthy
20-12-17, 22:24
Absolutely not. I'll repeat. No one who is half Lombard and half German is going to get these scores. They'd be more "northern", i.e. higher northern Europe.

I agree. I’m half Belgian (similar to southwest Germany) and half French Canadian which is an average of French HGDP according to FTDNA. On Gencove I get 68% Northern and Central, 27% Southwestern, and 5% Eastern Med.

Based on his scores his German is probably more northern Italian or his Italian does not have much of a Lombard influence.

alexfritz
20-12-17, 23:16
I agree. I’m half Belgian (similar to southwest Germany) and half French Canadian which is an average of French HGDP according to FTDNA. On Gencove I get 68% Northern and Central, 27% Southwestern, and 5% Eastern Med.

Based on his scores his German is probably more northern Italian or his Italian does not have much of a Lombard influence.

as i mentioned on the other thread, #33 is the concatenated b37[FTDNA] build means it incl the x-chr which is solely from the mother but not sure if gencove factors that in however, yet when i send in initially just the autosome build b36_autos[FTDNA] without the x-chr that was my result;
9554

not sure what academic papers you have seen but lombards(bergamo etc.) are fully south european close to tuscans, greeks and albanians (not french or cen_eu) and with a strong affinity towards modern sardinians which is neolithic based; for example in recent papers like haber et al the most shared alleles/drift with sidon_ba were modern sardinians and north_italians (neolithic based) or fregel et al TOR is more similar to Middle Neolithic/Chalcolithic Europeans, and, among modern populations, to populations from Spain, North Italy and Sardinia plus other such informative passages from other academic papers; and since this is the manner in which results are discussed, angela who is fully north italian only scored 8% NC but since she is not from lombardia i doubt she shared much drift with the sardinians/(orig.pop) in Geneplaza K25 (judging by her pardoned surprise) yet that is something someone from lombardia specifically ought to have (academic based); in several GEDmatch calculators a criteria to tell north_italian from south_italian is actually by east-med / west_med proportions not so much by other proportions;

davef
20-12-17, 23:36
Not sure what you mean, Alex. I don't think Sardinians are close to any other population, and north Italians and Greeks aren't terribly close either.

mwauthy
20-12-17, 23:58
as i mentioned on the other thread, #33 is the concatenated b37[FTDNA] build means it incl the x-chr which is solely from the mother but not sure if gencove factors that in however, yet when i send in initially just the autosome build b36_autos[FTDNA] without the x-chr that was my result;
9554

not sure what academic papers you have seen but lombards(bergamo etc.) are fully south european close to tuscans, greeks and albanians (not french or cen_eu) and with a strong affinity towards modern sardinians which is neolithic based; for example in recent papers like haber et al the most shared alleles/drift with sidon_ba were modern sardinians and north_italians (neolithic based) or fregel et al TOR is more similar to Middle Neolithic/Chalcolithic Europeans, and, among modern populations, to populations from Spain, North Italy and Sardinia plus other such informative passages from other academic papers; and since this is the manner in which results are discussed, angela who is fully north italian only scored 8% NC but since she is not from lombardia i doubt she shared much drift with the sardinians/(orig.pop) in Geneplaza K25 (judging by her pardoned surprise) yet that is something someone from lombardia specifically ought to have (academic based); in several GEDmatch calculators a criteria to tell north_italian from south_italian is actually by east-med / west_med proportions not so much by other proportions;

Thanks for the info! I always assumed northern Italians would plot closer to Central Europeans. You score 40% Southeastern Europe on FTDNA. I always assumed there was a large farming component in the West and Central Europe category and that Southeastern was more Mediterranean than Northern Italian. I wonder what Angela scored on FTDNA if she ever took that specific test.

My mom is French Canadian and scored 99% West and Central Europe and French people have a significant Neolithic farming component.

Angela
21-12-17, 02:05
Thanks for the info! I always assumed northern Italians would plot closer to Central Europeans. You score 40% Southeastern Europe on FTDNA. I always assumed there was a large farming component in the West and Central Europe category and that Southeastern was more Mediterranean than Northern Italian. I wonder what Angela scored on FTDNA if she ever took that specific test.

My mom is French Canadian and scored 99% West and Central Europe and French people have a significant Neolithic farming component.

I never took FTDNA.

On some calculators I get 20% Northwestern European.

@alexfritz,
I'm not 100% Northern Italian, unless you consider eastern Ligurians (1/4) and northwestern Tuscans (1/4) Northern Italian.

On gedmatch calculators I plot right between Bergamo and Firenze, which makes perfect sense.

I'm quite different from someone from the Veneto or Bergamo, but I'm also very different from anyone south of Tuscany.

Pratt
21-12-17, 02:23
in several GEDmatch calculators a criteria to tell north_italian from south_italian is actually by east-med / west_med proportions not so much by other proportions;

Not really.

alexfritz
21-12-17, 02:40
Thanks for the info! I always assumed northern Italians would plot closer to Central Europeans. You score 40% Southeastern Europe on FTDNA. I always assumed there was a large farming component in the West and Central Europe category and that Southeastern was more Mediterranean than Northern Italian. I wonder what Angela scored on FTDNA if she ever took that specific test.

My mom is French Canadian and scored 99% West and Central Europe and French people have a significant Neolithic farming component.

cant find the link anymore, but someone gave me a link as to what the myOrigins is modeled on and if i remember it correctly than West and Central incl french-swiss and southeastern the sardinians; and if there is one pattern i have noticed by my results is that i am wherever the sardinians are in gencove southwestern in myorigins southeastern in genographic southern and the K25/K29 on geneplaza underlined it all just nicely; i assumed that solely based on the academic data, guess not the wrongest reference to go by; not sure about french but with sardinian you can be certain that EEF goes hand in hand;


Not really.

yes, really is the case;

Angela
21-12-17, 03:11
Eurogenes calculators are terrible for Italians, so I wouldn't rely on them for any trends...

That whole "East Med" category is bogus as far as I'm concerned, unless by East Med is meant Greece and the Aegean islands. If people want to look at SW Asian, then look at that. If they want to look at Caucasus, then look at that. In terms of Caucasus, Italians, even Southern Italians, don't have any more Caucasus than Balkan people. Southern Italians have a bit more SW Asian.

The only calculator that shows a true percentage for West Med versus East Med in a European context is, imo, the ancient dna one done by Kurd. The "eastern" farmers stretched from the Aegean/western Turkey into Central Europe. The SWestern one of which I got a lot is based on a lot of late Neolithic Spanish samples and so has at least 20% WHG. Both Cardial/Impressa and the mainland Balkans farmers started out in the same general area so I doubt that they were all that different, although it would be nice to have a full genome analysis of a Cardial/ Impressa sample. By the late Neolithic, I think some additional gene flow from perhaps Anatolia into the Balkans and then perhaps from there into Italy, brought some additional CHG, but some analyses show a bit of CHG even in LBK.

That Eurogenes terminology obfuscates more than it illuminates, imo.

davef
21-12-17, 03:56
Alex says:
"and if there is one pattern i have noticed by my results is that i am wherever the sardinians are in gencove southwestern in myorigins southeastern in genographic southern and the K25/K29 on geneplaza underlined it all just nicely"

sorry, I didn't get any of that

alexfritz
21-12-17, 04:00
Alex says:
"and if there is one pattern i have noticed by my results is that i am wherever the sardinians are in gencove southwestern in myorigins southeastern in genographic southern and the K25/K29 on geneplaza underlined it all just nicely"

sorry, I didn't get any of that

what exactly?

davef
21-12-17, 05:03
what exactly?
That entire quote. I can't decipher it.

alexfritz
21-12-17, 05:27
That entire quote. I can't decipher it.

just so there are no misunderstandings i am not complaining/questioning any of my results in fact being half central european and half south european i think all my results are spot on; i like them a lot;

as for the quote, if i understood it correctly than clusters are based on populations and if such a cluster incl the sardinians my southern half always scores/corresponds with that cluster as should be for north italian (academic based); a quick example there was a calculator called K25 and on that calculator was a cluster called sardinian-sicilian, now not many people scored it and app a lot complained not scoring it so the creator in bold fat letters (mwauthy knows what im talking about) explained that the reason not many score it is because the sardinians are isolated yet if someone did score it it would indicate shared drift with the isolated sardinians and even claimed that that is synonymous with EEF; i scored proper 27% of that sardinian drift (only half north italian!) so it is just a pattern i have noticed by my own results which corresponds to academic results;

seems as if anthrogenica is knocked out again, when it works i link the link of the creator to you via PM;

Sile
21-12-17, 07:01
47% Southwestern Europe
40% Northern and Central Europe
5% Eastern Mediterranean
5% Northeast Europe
3% Iran Caucasus Anatolia

https://i.imgur.com/DKBO8EO.png

Gencove is like many who note these areas

SW Europe = iberia , southern france and NW Italy representing the Gallic influence

NW Europe = north france, Britain, swiss , austria, lombard and NE italy, denmark and germany

Greece and albania with southern Italy

Others place bulgari, romania and hungaria seperate ...gencove does not

the top 3 I have seen on many occasions by many companies

Angela
21-12-17, 07:13
It was a Sardinian/Sicilian cluster. IT WAS NOT JUST SARDINIAN.

It could as well mean closeness to Sicilians.

alexfritz
21-12-17, 08:07
It was a Sardinian/Sicilian cluster. IT WAS NOT JUST SARDINIAN.

It could as well mean closeness to Sicilians.

def not in my case i score 0.0% sicilian k29 yet 27% sardinian-sicilian k25 (the pattern); when anthrogenica works again ill link to the creators explanation #533 of sardinian-sicilian being sardinian 'original population drift' meaning you can read it first hand (from the creator);

@Sile
for the two clusters you mentioned gencove writes the following

Southwestern Europe

Reference populations: Southern French, Spanish, Basque, and Sardinian.

Northern and Central Europe

Reference populations: Norwegian, Orcadian, Scottish, British, Icelandic.

mwauthy
21-12-17, 20:38
just so there are no misunderstandings i am not complaining/questioning any of my results in fact being half central european and half south european i think all my results are spot on; i like them a lot;

as for the quote, if i understood it correctly than clusters are based on populations and if such a cluster incl the sardinians my southern half always scores/corresponds with that cluster as should be for north italian (academic based); a quick example there was a calculator called K25 and on that calculator was a cluster called sardinian-sicilian, now not many people scored it and app a lot complained not scoring it so the creator in bold fat letters (mwauthy knows what im talking about) explained that the reason not many score it is because the sardinians are isolated yet if someone did score it it would indicate shared drift with the isolated sardinians and even claimed that that is synonymous with EEF; i scored proper 27% of that sardinian drift (only half north italian!) so it is just a pattern i have noticed by my own results which corresponds to academic results;

seems as if anthrogenica is knocked out again, when it works i link the link of the creator to you via PM;

So according to Gencove northern Italians and Sardinians are more similar to Iberians, but on FTDNA they are more similar to Greeks? Which is more accurate according to the academic papers?

davef
21-12-17, 20:48
Don't have much time to browse, but I doubt north Italians are similar to any Greeks at all, Greece is way closer to south Italy and Albania

mwauthy
21-12-17, 20:59
cant find the link anymore, but someone gave me a link as to what the myOrigins is modeled on and if i remember it correctly than West and Central incl french-swiss and southeastern the sardinians; and if there is one pattern i have noticed by my results is that i am wherever the sardinians are in gencove southwestern in myorigins southeastern in genographic southern and the K25/K29 on geneplaza underlined it all just nicely; i assumed that solely based on the academic data, guess not the wrongest reference to go by; not sure about french but with sardinian you can be certain that EEF goes hand in hand;



yes, really is the case;

I’m curious as to what exactly the “Iberian” category entails on FTDNA? It used to be part of “Southern Europe” prior to 2.0. It doesn’t seem to include southwest France since my mom and I received 0% Iberian. It doesn’t seem to include Sardinia either. If you could find that link your friend sent you regarding the FTDNA regions that would be awesome!

Sile
21-12-17, 21:06
def not in my case i score 0.0% sicilian k29 yet 27% sardinian-sicilian k25 (the pattern); when anthrogenica works again ill link to the creators explanation #533 of sardinian-sicilian being sardinian 'original population drift' meaning you can read it first hand (from the creator);

@Sile
for the two clusters you mentioned gencove writes the following

Southwestern Europe

Reference populations: Southern French, Spanish, Basque, and Sardinian.

Northern and Central Europe

Reference populations: Norwegian, Orcadian, Scottish, British, Icelandic.



my Gencove


37% Eastern Mediterranean
37% Northern and Central Europe
23% Southwestern Europe


I have never had more than 2% iberian on any company.........but have had a high percentage of French-basque from many companies ............maybe my 23% SW Europe is that they used all the border people of Spain and France as iberians

Pax Augusta
21-12-17, 21:17
So according to Gencove northern Italians and Sardinians are more similar to Iberians, but on FTDNA they are more similar to Greeks? Which is more accurate according to the academic papers?

FTDNA is quite ridiculous.

Angela
21-12-17, 21:51
According to Gencove, northeastern and north central Italy seem to be Northern and Central European like, and northwestern Italy all the way down through Tuscany is Iberian like. In central Lombardia there's some overlap. That's going by the map upthread.

I've given up on results from these companies. I'm totally typical for where I come from, always placing midway between Bergamo and TSI samples. Academic studies of ancient dna tell me and will tell me how I got that way.

These companies are making all sorts of subjective decisions as to which clusters they'll use, and therefore the results are all different. It just confuses most people. That said, some tend to be better than others. I never tested at FTDNA because, indeed, as Pax says, they seemed ridiculous. For goodness sakes, in the beginning, and even now for all I know they use Ashkenazim, with their perhaps 60% European as a reference for the Middle East. McDonald used to do the same. That is completely NUTS.

The only thing I'll give gencove is that it duplicates all the Mid-to-late Spanish like Neolithic and Chalcolithic that I found through Kurd's ancient samples calculator. That doesn't mean a migration from Spain to Northwest Italy, for anyone who is wondering.

Jovialis
21-12-17, 22:08
According to Gencove, northeastern and north central Italy seem to be Northern and Central European like, and northwestern Italy all the way down through Tuscany is Iberian like. In central Lombardia there's some overlap. That's going by the map upthread.
I've given up on results from these companies. I'm totally typical for where I come from, always placing midway between Bergamo and TSI samples. Academic studies of ancient dna tell me and will tell me how I got that way.
These companies are making all sorts of subjective decisions as to which clusters they'll use, and therefore the results are all different. It just confuses most people. That said, some tend to be better than others. I never tested at FTDNA because, indeed, as Pax says, they seemed ridiculous. For goodness sakes, in the beginning, and even now for all I know they use Ashkenazim, with their perhaps 60% European as a reference for the Middle East. McDonald used to do the same. That is completely NUTS.
The only thing I'll give gencove is that it duplicates all the Mid-to-late Spanish like Neolithic and Chalcolithic that I found through Kurd's ancient samples calculator. That doesn't mean a migration from Spain to Northwest Italy, for anyone who is wondering.


https://www.youtube.com/watch?v=2J5n8kzpbCs

Absolutely true, even in this dumb commercial they perpetuate this misconception. This guy says he thought he was German, which he definitely is since he apparently knows his lineage. Yet, he actually thinks he's Scottish, and not German at all just because the test uses that population as a reference for populations that overlap between West Germany, and Scotland. I think some of this is for marketing purposes, so they can sell to people who may be surprised by their results, by finding something they think they didn't know about. At least that's what I'm getting from this commercial. The geneticists behind the test certainly know the truth behind what those reference populations actually imply. I get that its a business, and they want to make money, but this is not useful to people unless they are more adept in understanding how to gauge the results. Certain limitations on how to compose the algorithm is one thing. But to sell the results from these limitations, as they do in this commercial, as some kind of fact is just unethical. People should be informed that these reference populations don't necessarily mean that they come from these areas.

davef
21-12-17, 22:23
https://www.youtube.com/watch?v=2J5n8kzpbCs

Absolutely true, even in this dumb commercial they perpetuate this misconception. This guy says he thought he was German, which he definitely is since he apparently knows his lineage. Yet, he actually thinks he's Scottish, and not German at all just because the test uses that population as a reference for populations that overlap between West Germany, and Scotland. I think some of this is for marketing purposes, so they can sell to people who may be surprised by their results, by finding something they think they didn't know about. At least that's what I'm getting from this commercial. The geneticists behind the test certainly know the truth behind what those reference populations actually imply. I get that its a business, and they want to make money, but this is not useful to people unless they are more adept in understanding how to gauge the results.

Thats committing fraud on their part, unless they try to explain why this would happen in a simple but informative way. I doubt Bobby Joe has any desire to dig deep through a textbook on genetics.

mwauthy
21-12-17, 22:54
Thats committing fraud on their part, unless they try to explain why this would happen in a simple but informative way. I doubt Bobby Joe has any desire to dig deep through a textbook on genetics.
Yeah! None of these tests can tell you where exactly your ancestors were born and should not be used to override a paper trail like in that commercial. All these tests can do is show certain allele frequencies based on subjective regional groupings. For the average layman unaware of that they are going to waste a lot of time looking for a “lost ancestor” in their paper trail that doesn’t exist. For example, being French I get around 25% Iberian on many calculators. If I knew nothing about the genetic history of these regions I would be looking for a Spanish grand parent in my family tree.

It’s also misleading to call it an “ethnicity” estimate. A person’s ethnicity has no correlation with their alleles. For example, my ethnicity is United States. Ethnicities can also be generational and who knows what someone’s ethnicity was during the Bronze Age. These tests should be called allele frequency estimators.

alexfritz
22-12-17, 00:43
So according to Gencove northern Italians and Sardinians are more similar to Iberians, but on FTDNA they are more similar to Greeks? Which is more accurate according to the academic papers?

no, you skipped the part about sardinians actually being a part of that cluster #50 at gencove; the other south cluster is a wild mash up from bulgarians > sicilians so not sure what the artist is trying to tell us with that, but it could correspond to gedmatch gradient of west_med / east_med proportions in confirming it; there is no such thing as an/one italian cluster thus the calc-creators have to decide between the two main ones north_it/tuscan or south_it/sicilian it seems as if 23andme went for the latter yet myHeritage for the former; in toto diff calcs diff clusters, hence you have to look at the ref pops used in each cluster of each calc and academic papers/info will than explain/reveal the dynamics behind it;


I’m curious as to what exactly the “Iberian” category entails on FTDNA? It used to be part of “Southern Europe” prior to 2.0. It doesn’t seem to include southwest France since my mom and I received 0% Iberian. It doesn’t seem to include Sardinia either. If you could find that link your friend sent you regarding the FTDNA regions that would be awesome!

dont know i dont score it, the sardinians are part of the 'southeastern';
it seems i lost the link, you have to go onto the FTDNA forum (or gedmatch?) and ask for a spreadsheet, i did that back in june maybe they still have it; there is also the 'LearningCenter' on FTDNA (but it is not the link)
https://www.familytreedna.com/learn/user-guide/family-finder-myftdna/myorigins-population-clusters/

Jovialis
22-12-17, 01:10
no, you skipped the part about sardinians actually being a part of that cluster #50 at gencove; the other south cluster is a wild mash up from bulgarians > sicilians so not sure what the artist is trying to tell us with that, but it could correspond to gedmatch gradient of west_med / east_med proportions in confirming it; there is no such thing as an/one italian cluster thus the calc-creators have to decide between the two main ones north_it/tuscan or south_it/sicilian it seems as if 23andme went for the latter yet myHeritage for the former; in toto diff calcs diff clusters, hence you have to look at the ref pops used in each cluster of each calc and academic papers/info will than explain/reveal the dynamics behind it;
dont know i dont score it, the sardinians are part of the 'southeastern';
it seems i lost the link, you have to go onto the FTDNA forum (or gedmatch?) and ask for a spreadsheet, i did that back in june maybe they still have it; there is also the 'LearningCenter' on FTDNA (but it is not the link)
https://www.familytreedna.com/learn/user-guide/family-finder-myftdna/myorigins-population-clusters/

What happened to the fun-fact?

alexfritz
22-12-17, 01:12
What happened to the fun-fact?

the myheritage results? anthrogenica is still down no link possible to the sikeliot posts, when it works again ill link it all; would make more sense in that way (with a link attached);

Jovialis
22-12-17, 01:19
Why did you omit it from your post? Just curious.

alexfritz
22-12-17, 01:29
Anthrogenica was taken down by his post??????

I thought you said you got the information courtesy of him.

??? what ?
well he didnt give them to me he posted them public but yes courtesy of sikeliot you can find a lot of myHeritage results south italians and sicilians in that thread on anthrogenica; revealing a diff dynamic in this calc from other calcs despite same populations, now is that more fun with or without a link?

Jovialis
22-12-17, 01:32
??? what ?
well he didnt give them to me he posted them public but yes courtesy of sikeliot you can find a lot of myHeritage results south italians and sicilians in that thread on anathrogenica; revealing a diff dynamic in this calc from other calcs despite same populations, now is that more fun with or without a link?

Well if you understand how these calculators are made, the centrum for what is "Italian" can be different from test to test. Myheritage claiming that one is more than another means absolutely nothing. Fun right?

alexfritz
22-12-17, 01:37
Well if you understand how these calculators are made, the centrum for what is "Italian" can be different from test to test. Myheritage claiming that one is more than another means absolutely nothing. Fun right?

exactly, and exactly what i wrote #60; myHeritage (judging by the results/ but no link) defined their 'italian' based on the north_it/tuscan cluster hence that is what that means, and why so many south italians and sicilians do not score it in large amounts several in fact as dismal 10% and maybe now youll understand that all of those fun-facts would make so much more sense with the according link attached, but the link does not work;

Jovialis
22-12-17, 01:40
exactly, and exactly what i wrote #60; myHeritage (judging by the results/ but no link) defined their 'italian' based on the north_it/tuscan cluster hence that is what that means, and why so many south italians and sicilians do not score it in large amounts several in fact as dismal 10% and maybe now youll understand that all of those fun-facts would make so much more sense with the according link attached, but the link does not work;

Probably better off not working, since its a garbage assessment.

10% that's a joke.

alexfritz
22-12-17, 01:52
Probably better off not working, since its a garbage assessment.

10% that's a joke.

well now you have your answer as to why i omitted it, because there is no point in discussing numbers from a link that does not work meaning if you had the link yourself you could make your own assessment and when it works i will give you that link, basically just results sikeliot posted (i think he even created a spreadsheet);

Pax Augusta
22-12-17, 02:05
will give you that link, basically just results sikeliot posted (i think he even created a spreadsheet);

Ok, now everything is clear.

mwauthy
22-12-17, 02:33
exactly, and exactly what i wrote #60; myHeritage (judging by the results/ but no link) defined their 'italian' based on the north_it/tuscan cluster hence that is what that means, and why so many south italians and sicilians do not score it in large amounts several in fact as dismal 10% and maybe now youll understand that all of those fun-facts would make so much more sense with the according link attached, but the link does not work;
DNA Land has its own Sardinian category that’s separate from northern Italy, Sicily, and Southwestern Europe. I’m curious as to what your DNA Land scores are?

I think K29 lacked a Northern Italian cluster because my Dad gets 27% South/Central European on DNA Land and 21% Italian on Myheritage, yet on K29 he gets Southwestern Europe instead.

Angela
22-12-17, 02:39
@Pax,

It certainly is.

If people don't know enough by now to realize that everything depends on where the centrum is placed, i.e. a somewhat subjective decision, and that therefore judgments like who is more "Italian", or "French" or anything else shouldn't even be a topic of discussion, then they shouldn't be holding forth on matters genetic. The judgments of anyone who does that sort of thing should be automatically discounted both on that topic and anything else.

alexfritz
22-12-17, 02:55
DNA Land has its own Sardinian category that’s separate from northern Italy, Sicily, and Southwestern Europe. I’m curious as to what your DNA Land scores are?

I think K29 lacked a Northern Italian cluster because my Dad gets 27% South/Central European on DNA Land and 21% Italian on Myheritage, yet on K29 he gets Southwestern Europe instead.

practically diffused between the two clusters east med(levantine) and west med(south europe)+ sardinian on their own;
9559

at myHeritage i get ~33% italian and 7.9% greek and at K29 it is all 37% greek-albanian but these are two different calculators with different definitions and clusters and kurd explained that if you remove the cluster which has the most dominant allele sharing which was sardinian-siclian in k25 you will be assigned to the second most dominant obv greek-albanian k29;

PS: do you know what happened to anthrogenica? did get hacked again?

mwauthy
22-12-17, 03:06
practically diffused between the two clusters east med(levantine) and west med(south europe)+ sardinian on their own;
9559
at myHeritage i get ~33% italian and 7.9% greek and at K29 it is all 37% greek-albanian but these are two different calculators with different definitions and clusters and kurd explained that if you remove the cluster which has the most dominant allele sharing which was sardinian-siclian in k25 you will be assigned to the second most dominant obv greek-albanian k29;
PS: do you know what happened to anthrogenica? did get hacked again?
That’s interesting because you received almost an equal amount from all 4 regions in the Italian/Mediterranean area and 0% Southwestern. I’m starting to lean towards DNA Land being one of the more reputable calculators.

Interesting that your northern Italian got assigned to Greek on k29 and my dad’s northern Italian got assigned to Southwestern.

I have no idea about Anthrogenica. It’s just a blank white screen on my phone when I try to access the website.

alexfritz
22-12-17, 04:50
That’s interesting because you received almost an equal amount from all 4 regions in the Italian/Mediterranean area and 0% Southwestern. I’m starting to lean towards DNA Land being one of the more reputable calculators.

Interesting that your northern Italian got assigned to Greek on k29 and my dad’s northern Italian got assigned to Southwestern.

I have no idea about Anthrogenica. It’s just a blank white screen on my phone when I try to access the website.

fully agree i think DNA.Land might even be the most detailed for south european ancestry yet the way i see it i do view all calcs (i tested with at least) as equal reputable; maybe its a shift between CHG/WHG which tilts the north_italian either side? though in my case it was first mostly assigned to sardinian-sicilian k25 and once the sardinians were removed it than got further assigned to greek-albanian k29 and in that regard i am seeing something like this haak et al 5way mixture EEF/EN based;
9564

same here, anthrogenica is just blank though the activity stream reads an error code, not sure but i read in report that someones security software picked something up shortly before it blanked;

unorigional
22-12-17, 04:54
To throw in my two cents here, I had the impression that Gencove was trying to be better at detecting ancestry at the continental resolution. If you go through their website they have an affiliation with an effort to sequence more South Asians. So, they may be paring down the number of European components to allow more clusters elsewhere?

I actually took the plunge and submitted a sample for their own analysis. They are doing sparse sequencing as opposed to using the SNP specific chips. I'm curious to see the difference, but their low coverage sequencing hits way more bases on the DNA and Joe Pickrell makes what seems to me a good case for it (https://medium.com/the-seeq-blog/it-is-time-to-replace-genotyping-arrays-with-sequencing-73535efa66ed). I guess I'll find out in late Jan if it makes any difference at all.

In the mean-time, here is what I got from Gencove for my 23andMe v4 data:

9562

Aside from the Ashkenazi, which is nonesense and probably just a poorly chosen grouping, the rest is pretty consistent with what I have seen in other calculators and what I know of my ancestry.

As for the motivations of the big testing companies and the sometimes bizzaro advertisements they do, just consider this:

Ancestry is tied to the Mormons. They have an obsession with genealogy and genetic genealogy makes a lot of sense. Their end game is not figuring out your admixture. They want more people that can be linked together via DNA to build their massive family tree database even more.

23andMe: they just want your medical data because that is going to be hugely valuable to sell to pharma companies. After all, they sequence your genome for the health data regardless of whether you pay for it or not (fun fact, you can take their data and input it into other services such as promethease and get all the health data they would have sold you and more).

Can't speak to the motivation of the others, but I personally don't think the admixtures they give you are their real end goals.

(edited to clarify that result is from Gencove with my 23andMe data)

Angela
22-12-17, 05:09
To throw in my two cents here, I had the impression that Gencove was trying to be better at detecting ancestry at the continental resolution. If you go through their website they have an affiliation with an effort to sequence more South Asians. So, they may be paring down the number of European components to allow more clusters elsewhere?

I actually took the plunge and submitted a sample for their own analysis. They are doing sparse sequencing as opposed to using the SNP specific chips. I'm curious to see the difference, but their low coverage sequencing hits way more bases on the DNA and Joe Pickrell makes what seems to me a good case for it (https://medium.com/the-seeq-blog/it-is-time-to-replace-genotyping-arrays-with-sequencing-73535efa66ed). I guess I'll find out in late Jan if it makes any difference at all.

In the mean-time, here is what I got for my 23andMe v4 data:

9562

Aside from the Ashkenazi, which is nonesense and probably just a poorly chosen grouping, the rest is pretty consistent with what I have seen in other calculators and what I know of my ancestry.

As for the motivations of the big testing companies and the sometimes bizzaro advertisements they do, just consider this:

Ancestry is tied to the Mormons. They have an obsession with genealogy and genetic genealogy makes a lot of sense. Their end game is not figuring out your admixture. They want more people that can be linked together via DNA to build their massive family tree database even more.

23andMe: they just want your medical data because that is going to be hugely valuable to sell to pharma companies. After all, they sequence your genome for the health data regardless of whether you pay for it or not (fun fact, you can take their data and input it into other services such as promethease and get all the health data they would have sold you and more).

Can't speak to the motivation of the others, but I personally don't think the admixtures they give you are their real end goals.

You linked to gencove data, not 23andme.

mwauthy
22-12-17, 05:25
fully agree i think DNA.Land might even be the most detailed for south european ancestry yet the way i see it i do view all calcs (i tested with at least) as equal reputable; maybe its a shift between CHG/WHG which tilts the north_italian either side? though in my case it was first mostly assigned to sardinian-sicilian k25 and once the sardinians were removed it than got further assigned to greek-albanian k29 and in that regard i am seeing something like this haak et al 5way mixture EEF/EN based;
9564

same here, anthrogenica is just blank though the activity stream reads an error code, not sure but i read in report that someones security software picked something up shortly before it blanked;

That’s really interesting that you say WHG vs CHG might make the difference. I never thought about looking at a sub component to see which way the major component might sway. Makes sense! We like to use labels and words to simplify our categorizations but really all components are multilayered.

Similar to when someone says that category is “Celtic” or “Germanic” or “Roman.” Really it means nothing because there is variation in each label and each label is multilayered.

unorigional
22-12-17, 05:26
You linked to gencove data, not 23andme.

That's right. I meant that this was Gencove's results using my 23andMe data. I won't have the results from Gencove's sequencing for about another month. I'll post that once I have it.

Angela
22-12-17, 05:29
That's right. I meant that this was Gencove's results using my 23andMe data. I won't have the results from Gencove's sequencing for about another month. I'll post that once I have it.

Great. That will be interesting to see.

Sile
22-12-17, 07:07
DNA land are one of the only ones that confirmed a relative of which I have communicated many times and have linked between each of our family trees at about year 1780
Name:
Anna Bolzan
7 cousin
25 shared segments
137.14 Total shared length
BTW, can we load more kits through DNAland now?

.................................................. ........
My wife also confirmed a cousin via DNaland and family trees

Name:
Guy Barbato

11 cousin

17 shared segments

71.44 Total shared length

Sile
22-12-17, 19:22
my Gencove
37% Eastern Mediterranean
37% Northern and Central Europe
23% Southwestern Europe
I have never had more than 2% iberian on any company.........but have had a high percentage of French-basque from many companies ............maybe my 23% SW Europe is that they used all the border people of Spain and France as iberians
Big oops ....I stuffed up, I only uploaded my fathers data to Gencove and above is his.
I uploaded to get the free prometheuse report which arrived yesterday..........his ills and blood type match the results and he had issues with these but did not pass away due to them.
My father did have SW Europe and had matches with an area covering Girona and Perpignan , he also had many matches with Toulouse and Montpellier areas due to relatives moving there between 1875 and 1920

srdceleva
23-12-17, 14:58
my results. Seems pretty descent.
https://preview.ibb.co/jHD8Yk/gencove_results.png (https://ibb.co/kfN2tk)

Fathers results
https://preview.ibb.co/bYNADk/gencove_dad.png (https://ibb.co/b2VK65)

alexfritz
24-12-17, 10:07
@Angela

#533 K25 sardinian-sicilian and why sardinians were removed from that cluster in K29
https://anthrogenica.com/showthread.php?12817-New-K25-Admixture-Calculator-at-GenePlaza-com/page54
It seems that the enhanced drift In Sardinians, Who are part of the SC European references in the K 25 is suppressing South Central European scores in European testers

Although isolation of the Sardinians is a good thing in terms of preserving South Central European substructure for the past few thousand years and thus they can be seen as genetically unchanged representatives of south-central Europeans as they were a few thousand years ago while populations around them in mainland Europe were constantly changing. This is what makes them more “pure” south-central Europeans than other populations in my view, and was one of the main reasons for their inclusion into my K 25 SC European component, and this is definitely a pro .

The con for their inclusion is that there enhanced genetic drift due to isolation is suppressing South Central European scores in testers some

So far we have seen with testers is that some west Asians and a few select Europeans ( mainly Italians) have the highest South Central European scores, an indicator that those testers share the highest drift with south-central Europe’s original inhabitants

The problem is that this is creating a little confusion for some testers who are used to seeing more elevated South Central European scores and other tests.

At This point I am still considering the pros versus the cons of including Sardinians in the south central European component and will make an announcement when I reach a decision

@ Jovialis p.66#655
https://anthrogenica.com/showthread.php?9093-MyHeritage-DNA/page66

Jovialis
24-12-17, 15:31
I already told you I don't care to see it.

I already said it was nothing but garbage too.

davef
24-12-17, 15:44
Alex, ignore everything sikeliot posts.

mwauthy
25-12-17, 01:33
My wife is half Italian (Naples and Northern Italy). Here are her Southern Europe scores from various companies:

23andMe:
Italian: 32.5%
Iberian: 5.2%
Broadly Southern: 8.9%

DNA Land:
South/Central European: 33%
Southwestern European: 4.6%


FTDNA:
Iberian: 32%
Southeastern Europe: 14%
Sephardic: 14%

Is the FTDNA algorithm really that off or does this mean it’s just really hard to differentiate Southern European alleles?

davef
25-12-17, 01:55
My wife is half Italian (Naples and Northern Italy). Here are her Southern Europe scores from various companies:

23andMe:
Italian: 32.5%
Iberian: 5.2%
Broadly Southern: 8.9%

DNA Land:
South/Central European: 33%
Southwestern European: 4.6%


FTDNA:
Iberian: 32%
Southeastern Europe: 14%
Sephardic: 14%

Is the FTDNA algorithm really that off or does this mean it’s just really hard to differentiate Southern European alleles?
How did 32-33 percent Italian turn into 32 percent Iberian? Yeah FTDNA isn't looking any better. Plus the questionable 14 percent Sephardic....

mwauthy
25-12-17, 02:28
How did 32-33 percent Italian turn into 32 percent Iberian? Yeah FTDNA isn't looking any better. Plus the questionable 14 percent Sephardic....

The map for Sephardic is the Iberian Peninsula. So my half British/ half Italian wife is 46% Iberian Map according to FTDNA.

mwauthy
25-12-17, 02:33
@Angela

#533 K25 sardinian-sicilian and why sardinians were removed from that cluster in K29
https://anthrogenica.com/showthread.php?12817-New-K25-Admixture-Calculator-at-GenePlaza-com/page54
It seems that the enhanced drift In Sardinians, Who are part of the SC European references in the K 25 is suppressing South Central European scores in European testers

Although isolation of the Sardinians is a good thing in terms of preserving South Central European substructure for the past few thousand years and thus they can be seen as genetically unchanged representatives of south-central Europeans as they were a few thousand years ago while populations around them in mainland Europe were constantly changing. This is what makes them more “pure” south-central Europeans than other populations in my view, and was one of the main reasons for their inclusion into my K 25 SC European component, and this is definitely a pro .

The con for their inclusion is that there enhanced genetic drift due to isolation is suppressing South Central European scores in testers some

So far we have seen with testers is that some west Asians and a few select Europeans ( mainly Italians) have the highest South Central European scores, an indicator that those testers share the highest drift with south-central Europe’s original inhabitants

The problem is that this is creating a little confusion for some testers who are used to seeing more elevated South Central European scores and other tests.

At This point I am still considering the pros versus the cons of including Sardinians in the south central European component and will make an announcement when I reach a decision

@ Jovialis p.66#655
https://anthrogenica.com/showthread.php?9093-MyHeritage-DNA/page66

I don’t think removing the Sardinians helped the South Central Europeans at all for K29 because Sicily alone is not a good proxy. Central or Northern Italian samples might have been better. My dads 27% South Central on DNA Land became all Southwestern on k29.

davef
25-12-17, 03:18
The map for Sephardic is the Iberian Peninsula. So my half British/ half Italian wife is 46% Iberian Map according to FTDNA.
It's only highlighted to show where Sephardic Jews heavily settled, the Iberian peninsula. It wouldn't be reasonable to highlight anyplace else, and to add, Sephardic Jews aren't Iberian genetically.

Being 1/4 from Naples with the rest being 1/4 north Italian and 1/2 British shouldn't make you 46 percent Iberian, from what I can tell.

Angela
26-12-17, 02:52
If there's no Southern Central component centered on Italy, everything else is going to be off. You can't just have SW and SE Europe.

Tuscany seems the logical choice: Northern Italy is too Northern and Southern Italy/Sicily too Southern. There's a reason 1000 Genomes chose them as the reference for the Central Med.

Until this calculator has such a cluster it will be useless for Italians or even for Balkans people. Having Tuscany as the reference would make clear the genetic relationship that we see with the Balkans from PCAs.

I said from the beginning that it was clear that Sardinians are far too drifted to be used. Sicilians as a reference isn't all that much better.

I'm sorry to say it but I'm very disappointed in this calculator.

alexfritz
15-01-18, 17:28
gencove UPDATE! 01/2018 with new clusters

FTDNAb_37_concatenated / FTDNAb_37_autos
9637 9638

Stuvanè
15-01-18, 20:27
I'm pleased that a specific cluster has been inserted for Northern Italy. But now my results diverge considerably from the previous oneshttps://uploads.tapatalk-cdn.com/20180115/5c5e50f8480b45909219c2f505f5c826.jpg

Inviato dal mio SM-A300FU utilizzando Tapatalk

Angela
15-01-18, 21:31
I'm pleased that a specific cluster has been inserted for Northern Italy. But now my results diverge considerably from the previous oneshttps://uploads.tapatalk-cdn.com/20180115/5c5e50f8480b45909219c2f505f5c826.jpg

Inviato dal mio SM-A300FU utilizzando Tapatalk

I'm about to give up on all of these things except 23and me. How does that make any sense for someone 100 per cent from the Romagna?

Sile
16-01-18, 18:57
mine upgraded yesterday.
,
https://s20.postimg.org/tln7wpd19/my_gencove.jpg (https://postimg.org/image/42uvjoth5/)
.
starting to get closer to yfull data which has my snp dated as 1460BC.
Yfull matches for me relating to this gencove data
1x Hessian germany
3x walloon and lorraine france ...........these could be close relatives to each other
1x south tyrol italy
1x spain Pyrenees area Mediterranean side
.
note: central asia is 1% ..........other sites have me as 0.8% for it

Angela
16-01-18, 19:13
How did you people get your maps to post?

For what it's worth:



57%Northern Italy
29%Southwestern Europe
7%Northern and Central Europe
7%Anatolia, Caucasus, Iranian Plateau



It's not bad, but if I didn't know my identity, this wouldn't tell me. The Northern Europe numbers go down, and the Anatolia, Caucasus shows up. Have no idea what that means.

https://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1dahttps://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1da

Dibran
16-01-18, 19:41
23andme:

93% Eastern Mediterranean
7% South-West Europe

AncestryDNA:

78% Eastern Mediterranean
22% South-West European

Sile
16-01-18, 19:42
How did you people get your maps to post?

For what it's worth:



57%Northern Italy
29%Southwestern Europe
7%Northern and Central Europe
7%Anatolia, Caucasus, Iranian Plateau



It's not bad, but if I didn't know my identity, this wouldn't tell me. The Northern Europe numbers go down, and the Anatolia, Caucasus shows up. Have no idea what that means.

https://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1dahttps://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1da

i use postimage , its free as long as you do not go overboard

snip map, save, upload on postimage

Angela
16-01-18, 19:50
23andme:

93% Eastern Mediterranean
7% South-West Europe

AncestryDNA:

78% Eastern Mediterranean
22% South-West European


Here's a blogpost from someone else who got results all over the place:

https://gizmodo.com/how-dna-testing-botched-my-familys-heritage-and-probab-1820932637

"The CEO of GenCove, the company where I had uploaded my 23andMe data to get drastically different results, told me that even though he expects a fair amount of variability between algorithms, even he was surprised at how differently his company and 23andMe had interpreted my DNA data. He asked me to also upload my Ancestry data, and ran both data sets again after GenCove’s algorithm had been updated. The results were all over the map.“To be honest I’m a little confused about what’s going on,” CEO Joseph Pickrell told me.


Each testing company is looking at different alleles from different parts of the genome, and using different algorithms to crunch that data. (You can see a list of how company tests differ here (https://isogg.org/wiki/Autosomal_DNA_testing_comparison_chart).) It’s worth mentioning that genetics is also probabilistic: just because you have the gene, doesn’t mean you have the trait."

I don't understand your high SW Euro from one company. In my case, we're right next door, and we have some "Spanish" and "French" yDna.

One thing is consistent, almost no one ever gets "Italian" scores higher than mine. My father, old school nationalist that he was, would be very happy. :)

Dibran
16-01-18, 20:57
Here's a blogpost from someone else who got results all over the place:

https://gizmodo.com/how-dna-testing-botched-my-familys-heritage-and-probab-1820932637

"The CEO of GenCove, the company where I had uploaded my 23andMe data to get drastically different results, told me that even though he expects a fair amount of variability between algorithms, even he was surprised at how differently his company and 23andMe had interpreted my DNA data. He asked me to also upload my Ancestry data, and ran both data sets again after GenCove’s algorithm had been updated. The results were all over the map.“To be honest I’m a little confused about what’s going on,” CEO Joseph Pickrell told me.


Each testing company is looking at different alleles from different parts of the genome, and using different algorithms to crunch that data. (You can see a list of how company tests differ here (https://isogg.org/wiki/Autosomal_DNA_testing_comparison_chart).) It’s worth mentioning that genetics is also probabilistic: just because you have the gene, doesn’t mean you have the trait."

I don't understand your high SW Euro from one company. In my case, we're right next door, and we have some "Spanish" and "French" yDna.

One thing is consistent, almost no one ever gets "Italian" scores higher than mine. My father, old school nationalist that he was, would be very happy. :)

I usually score Italian in every calc in some form or another. 14 Italian over all with LivingDNA. I wonder what is so off with their algorithm?

In AncestryDNA for instance, their main site has be as 84% Albania/Greece/Turkey, 15% North-East Italy/Croatia/Bosnia and 1% Finnish(LOL). I am surprised by the huge South-West Euro for the Ancestry upload. My Fathers by comparison(with 23andme) is 88% Eastern Mediterranean and 12% South-West European.

Someone mentioned people with Sardinian like admixtures are scoring South-West Europe in Gencove. It would make sense, since theres an obvious(for whatever reason) shift towards Sardinia for my father and myself.

davef
16-01-18, 21:06
Why do these groups have populations that are far apart from each other? Like southwest Europe has Sardinians who are far from everyone else and east med goes from Cypriots to Bulgarians.

I kinda brought this up before but there are still some holes to patch up. Lets consider the south west euro group, does it just take everything that's common between Sardinians, Spaniards etc and deem this pattern of genes "Southwest European"?

AdeoF
16-01-18, 21:13
Now i got
Southwestern Europe 46%
Northern and Central Europe 28%
Northern Italy 12%
Eastern Mediterranean 11%

I still think it needs more work... Also Eastern Mediterranean doesn't make much sense??

Angela
16-01-18, 21:40
Now i got
Southwestern Europe 46%
Northern and Central Europe 28%
Northern Italy 12%
Eastern Mediterranean 11%

I still think it needs more work... Also Eastern Mediterranean doesn't make much sense??

Click to enlarge...

This is SE Europe:

9654
9659
This is Southwest Europe: it includes the southern half of France and Northwest Italy. Basically, I could just add this percentage to my North Italian.

9656
9660
This is northern Italy. It overlaps with North Central Europe, which is probably what brought those numbers down for me.
9661

9655

Maybe it is sort of right for me.

@Sile,
Thanks.

It works like any other Admixture program...it finds clusters. Some of the clusters overlap, which is indeed the reality.

Angela
16-01-18, 21:47
I usually score Italian in every calc in some form or another. 14 Italian over all with LivingDNA. I wonder what is so off with their algorithm?

In AncestryDNA for instance, their main site has be as 84% Albania/Greece/Turkey, 15% North-East Italy/Croatia/Bosnia and 1% Finnish(LOL). I am surprised by the huge South-West Euro for the Ancestry upload. My Fathers by comparison(with 23andme) is 88% Eastern Mediterranean and 12% South-West European.

Someone mentioned people with Sardinian like admixtures are scoring South-West Europe in Gencove. It would make sense, since theres an obvious(for whatever reason) shift towards Sardinia for my father and myself.

I don't think the problem with the Living DNA results necessarily has anything to do with their algorithm. They have uploaded a huge amount of samples from either the English or Irish Atlas, maybe both. The rest of their samples are an order of magnitude smaller. That is going to make everyone who isn't English or Irish look more "western".

Ordas
16-01-18, 21:54
My results on Gencove.

61%
Northern and Central Europe
26%
Eastern Mediterranean
7%
Northeast Europe
3%
Northern Italy
2%
Central Asia
1%
Southwestern Europe

I'm Hungarian from Transylvania

Gesendet von meinem SM-G903F mit Tapatalk

davef
16-01-18, 22:08
This is SE Europe:

9654

This is Southwest Europe: it includes the southern half of France and Northwest Italy. Basically, I could just add this percentage to my North Italian.

9656

This is northern Italy. It overlaps with North Central Europe, which is probably what brought those numbers down for me.


9655

Maybe it is sort of right for me.

@Sile,
Thanks.

It works like any other Admixture program...it finds clusters. Some of the clusters overlap, which is indeed the reality.
Well ok thanks, but grouping cypriots with Bulgarians and Sardinians with anyone else is insanity.

Angela
16-01-18, 22:26
Bulgarians and a lot of other groups in the Balkans have quite a bit more "Caucasus/Eastern Anatolia" than Northern Italians and Tuscans. They have more SW Asian than Northern Italians on some calculators, and just a bit less than Tuscans. I guess there was a migration of Samaritans to Bulgaria as well as Toscana. :)

So much disinformation and deliberately slanted data is published on internet sites that the reality that the Balkans, and not just southern Italy and Greece share a lot with places like Cyprus and eastern Anatolia is virtually unknown.

See, for example, the percentages for these clusters in Dodecad K12b:

https://docs.google.com/spreadsheets/d/1GWhNZcfTQ2hMSK9Ni1IqG7aXHB00SRE5L6ED2osPs9M/edit?hl=en_US&hl=en_US#gid=0

This is why Gencove, for instance, includes not only Southern Italy, but the Balkans and Greece in its "Eastern Mediterranean" cluster, and not just parts of Anatolia.

Boreas
16-01-18, 22:54
Well ok thanks, but grouping cypriots with Bulgarians and Sardinians with anyone else is insanity.


Insanity :laughing:

If you add a bit NorthEast Europe gens to Sardinian, you will have Bulgarian or add redsea admixture, you have Cypriot


Yeah similarity not as much as green-red apple. Mostly like Apple and Pear so on the less selective eyes, they are same.

--------------------------

As a 71 % Eastern Med in Gencove, I can say that they have killed Caucasia and NorthEast Europe.

davef
16-01-18, 23:05
Thanks Angela and Boreas, I'll think more into those posts later!

Regio X
17-01-18, 00:18
I'm pleased that a specific cluster has been inserted for Northern Italy. But now my results diverge considerably from the previous oneshttps://uploads.tapatalk-cdn.com/20180115/5c5e50f8480b45909219c2f505f5c826.jpg


mine upgraded yesterday.
,
https://s20.postimg.org/tln7wpd19/my_gencove.jpg (https://postimg.org/image/42uvjoth5/)
.
starting to get closer to yfull data which has my snp dated as 1460BC.
Yfull matches for me relating to this gencove data
1x Hessian germany
3x walloon and lorraine france ...........these could be close relatives to each other
1x south tyrol italy
1x spain Pyrenees area Mediterranean side
.
note: central asia is 1% ..........other sites have me as 0.8% for it


How did you people get your maps to post?
For what it's worth:

57%Northern Italy
29%Southwestern Europe
7%Northern and Central Europe
7%Anatolia, Caucasus, Iranian Plateau

It's not bad, but if I didn't know my identity, this wouldn't tell me. The Northern Europe numbers go down, and the Anatolia, Caucasus shows up. Have no idea what that means.
https://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1dahttps://ancestry.gencove.com/#/user/67869966-44ce-4ef8-9f9f-354f4e05f1da

Mine (FTDNA / 23andMe)

Northern and Central Europe: 44% / 43%
Eastern Mediterranean: 29% / 33%
Northern Italy: 23% / 13%
Southwestern Europe: 4% / 8%
Ashkenazi Jewish: 0% / 2%
Central Asia: 0% / 1%

Father (23andMe)

Northern Italy: 32%
Northern and Central Europe: 31%
Southwestern Europe: 16%
Eastern Mediterranean: 11%
Ashkenazi Jewish: 5%
Northern British Isles: 4%
Central Asia: 1%

Mother (23andMe)

Northern and Central Europe: 33%
Eastern Mediterranean: 28%
Northern Italy: 28%
Southwestern Europe: 11%

davef
17-01-18, 00:31
Mine (FTDNA / 23andMe)

Northern and Central Europe: 44% / 43%
Eastern Mediterranean: 29% / 33%
Northern Italy: 23% / 13%
Southwestern Europe: 4% / 8%
Ashkenazi Jewish: 0% / 2%
Central Asia: 0% / 1%

Father (23andMe)

Northern Italy: 32%
Northern and Central Europe: 31%
Southwestern Europe: 16%
Eastern Mediterranean: 11%
Ashkenazi Jewish: 5%
Northern British Isles: 4%
Central Asia: 1%

Mother (23andMe)

Northern and Central Europe: 33%
Eastern Mediterranean: 28%
Northern Italy: 28%
Southwestern Europe: 11%
Dang! North Italy as low as 13 percent!

If this doesnt get a reaction from Angela.. :grin::grin:

seriously, this thing is garbage

AdeoF
17-01-18, 00:52
This is SE Europe:

9654

This is Southwest Europe: it includes the southern half of France and Northwest Italy. Basically, I could just add this percentage to my North Italian.

9656

This is northern Italy. It overlaps with North Central Europe, which is probably what brought those numbers down for me.


9655

Maybe it is sort of right for me.

@Sile,
Thanks.

It works like any other Admixture program...it finds clusters. Some of the clusters overlap, which is indeed the reality.
I can't see the Attachments Angela

Salento
17-01-18, 01:01
Was 38% Southwestern Europe, Now is 4%, and 34% North Italy.

Angela
17-01-18, 01:25
If someone who had been adopted ran their data through this calculator, he/she would never figure it out.

I have said repeatedly that there is more genetic variation north of Rome than there is south of it, and I'm fully aware of the east/west structure in Italy, as well as the north/south one.

However, these companies are all marketing to the general population: therefore, you shouldn't need to have spent years studying population genetics to figure it out.

I'm done with this stuff.

Angela
17-01-18, 01:39
I can't see the Attachments Angela

Try it now, Adeo. Sorry about the difficulty.

Angela
17-01-18, 01:58
Dang! North Italy as low as 13 percent!

If this doesnt get a reaction from Angela.. :grin::grin:

seriously, this thing is garbage

Well, if you know a lot about Population Genetics, and how Admixture works, and take a careful look at their clusters you can figure it out.

Since SW Europe and NW Italy overlap you can add that, and since North Central Europe overlaps Northern Italy you could add some of that, and since half of Italy is within Eastern Med you could add that. That would be the "Italian" total. Get the idea?

The point, as I said upthread, is that I only know these things because I've been studying population genetics for eight years. The calculators are being used by people who don't have a clue about any of these things, and all it's going to do is confuse them, as well as bringing population genetics into disrepute.

Sile
17-01-18, 07:11
Mine (FTDNA / 23andMe)
Northern and Central Europe: 44% / 43%
Eastern Mediterranean: 29% / 33%
Northern Italy: 23% / 13%
Southwestern Europe: 4% / 8%
Ashkenazi Jewish: 0% / 2%
Central Asia: 0% / 1%
Father (23andMe)
Northern Italy: 32%
Northern and Central Europe: 31%
Southwestern Europe: 16%
Eastern Mediterranean: 11%
Ashkenazi Jewish: 5%
Northern British Isles: 4%
Central Asia: 1%
Mother (23andMe)
Northern and Central Europe: 33%
Eastern Mediterranean: 28%
Northern Italy: 28%
Southwestern Europe: 11%
it only looks like 23andme pulls your north italy down to 13% ...ftdna has you higher.

it also looks like the more west you go from Veneto/friuli/trentino towards turin and Genoa the more higher the North Italian % becomes

davef
17-01-18, 07:54
If someone who had been adopted ran their data through this calculator, he/she would never figure it out.

I have said repeatedly that there is more genetic variation north of Rome than there is south of it, and I'm fully aware of the east/west structure in Italy, as well as the north/south one.

However, these companies are all marketing to the general population: therefore, you shouldn't need to have spent years studying population genetics to figure it out.

I'm done with this stuff.
Edit: sorry guys, I was wrong, we cool

Angela
17-01-18, 14:36
The east and west doesn't seem all that related, the west has stronger ties to Iberia and the east has stronger ties to Greece/Balkans going by what's in this thread.

That's a vast over-generalization, and illogical.

To say that there's some structure east/west doesn't mean they're "not all that related."

This is almost "Fire-Haired" type reasoning.

The north was not only subject to some differing migration flows, it was also fragmented politically, with resulting isolation from one another and some drift. The south was, for most of the last 1000 years, ruled as a political entity.

mwauthy
17-01-18, 15:39
Some close similarities between my Ancestry dot com results and this new update from Gencove. Same with my mother.

Ancestry:
Europe West 47%
Iberian Peninsula 25%
Scandinavia 13%
Great Britain 6%

Gencove:
Northern and Central Europe 46%
Southwestern Europe 27%
Scandinavia 12%
Northern British Isles 6%

Mom Ancestry:
Iberian Peninsula: 32%
Great Britain: 23%
Scandinavia: 17%

Mom Gencove:
Southwestern Europe: 34%
Northern British Isles: 23%
Scandinavia: 17%

davef
17-01-18, 15:45
That's a vast over-generalization, and illogical.

To say that there's some structure east/west doesn't mean they're "not all that related."

This is almost "Fire-Haired" type reasoning.

The north was not only subject to some differing migration flows, it was also fragmented politically, with resulting isolation from one another and some drift. The south was, for most of the last 1000 years, ruled as a political entity.
Ok that's cool, I was just surprised at how different the scores were. Those differences threw me off.

alexfritz
17-01-18, 17:10
The east and west doesn't seem all that related, the west has stronger ties to Iberia and the east has stronger ties to Greece/Balkans going by what's in this thread.

relatedness is not determined by admixture analysis but by IBD analysis (segments), the admixture analysis only determines your broad regional ancestry by the given defined clusters; the more detailed the better the insight yet it only goes so far as the overlap; of course north_italy has a lot of overlap with other clusters hence it was not used in GenePlaza K25/K29 but the cool thing by gencove update is here you can now see where and to what extent the overlap was in the previous version and get some insights from that; the way i see it is that as already suspected a lot was overlapped in south_western and partially (previous)north_central;

PS: if you still get a lot of south_western despite a north_it cluster now, than you truly do have some either Iberian or Sardinian type ancestry this in turn i reckon can be filtered out via comparison to other calcs with such defined clusters;

mwauthy
17-01-18, 17:34
relatedness is not determined by admixture analysis but by IBD analysis (segments), the admixture analysis only determines your broad regional ancestry by the given defined clusters; the more detailed the better the insight yet it only goes so far as the overlap; of course north_italy has a lot of overlap with other clusters hence it was not used in GenePlaza K25/K29 but the cool thing by gencove update is here you can now see where and to what extent the overlap was in the previous version and get some insights from that; the way i see it is that as already suspected a lot was overlapped in south_western and partially (previous)north_central;

PS: if you still get a lot of south_western despite a north_it cluster now, than you truly do have some either Iberian or Sardinian ancestry this in turn i reckon can be filtered out via comparison to other calcs with such defined clusters;

Very true! My dad’s 21-27% northern Italian was included into Europe West on Ancestry and into Southwestern on K25/29 and old Gencove.

Angela
17-01-18, 17:46
^^That's another unwarranted conclusion. It seems to be the day for it.

I still get quite a bit of Southwestern European despite the fact that I get a lot of North Italian. It has nothing to do with any migration from Sardinia or Iberia.

First of all, there was never, until very recently indeed, any migration from Sardinia to the mainland. It all went the other way.

There is also no migration from Iberia to Liguria or Emilia.

In my own particular case, I can trace my ancestry back to the middle of the 1500s, and on some lines back to the 1100s and 1200s, and there is no such gene flow. In my father's case, in particular, all indications are that they were up in the northern Apennines by around 1000 AD.

Part of it, in the case of Liguria or even nearby areas in Emilia, may be attributable to Gallic migrations of the first millennium BC. There's a reason these people were called Celt-Ligurians. It is clear from Gencove's map that South Western Europe doesn't just include Iberia. It also applies to the southern half of France.

In addition, I would propose that most of northern Italy was once probably "Ligurian like". One of the differences is differential impact of the Lombard invasions. Yes, every hill in my valley is crowned with a Lombard castle. However, they entered through the northeastern corridor of Italy, and the density of settlement is strongest there. Also, the Lombards seem to have incorporated other men and women as they moved, so they may have been different by the time they incorporated some parts of Northwestern Italy.

Drift due to long separation would also have added to the structure.

In this calculator North Italy is centered on Northwestern Italy plus Tuscany.

As for "relatedness", we all know or should have known that Davef meant genetic similarity.

davef
17-01-18, 17:51
Thanks Angela, that is exactly what I meant by relatedness.

alexfritz
17-01-18, 18:14
^^That's another unwarranted conclusion. It seems to be the day for it.

I still get quite a bit of Southwestern European despite the fact that I get a lot of North Italian. It has nothing to do with any migration from Sardinia or Iberia.

First of all, there was never, until very recently indeed, any migration from Sardinia to the mainland. It all went the other way.

There is also no migration from Iberia to Liguria or Emilia.

In my own particular case, I can trace my ancestry back to the middle of the 1500s, and on some lines back to the 1100s and 1200s, and there is no such gene flow. In my father's case, in particular, all indications are that they were up in the northern Apennines by around 1000 AD.

Part of it, in the case of Liguria or even nearby areas in Emilia, may be attributable to Gallic migrations of the first millennium BC. There's a reason these people were called Celt-Ligurians. It is clear from Gencove's map that South Western Europe doesn't just include Iberia. It also applies to the southern half of France.

In addition, I would propose that most of northern Italy was once probably "Ligurian like". One of the differences is differential impact of the Lombard invasions. Yes, every hill in my valley is crowned with a Lombard castle. However, they entered through the northeastern corridor of Italy, and the density of settlement is strongest there. Also, the Lombards seem to have incorporated other men and women as they moved, so they may have been different by the time they incorporated some parts of Northwestern Italy.

Drift due to long separation would also have added to the structure.

In this calculator North Italy is centered on Northwestern Italy plus Tuscany.

As for "relatedness", we all know or should have known that Davef meant genetic similarity.

that is for sure, and that is how much missing a word or using a false word distorts (as related meant related to me) #122 should have been 'type ancestry' or 'substrate type' ancestry and not implying a direct migration ancestry (though can never be ruled out either); i think the listed populations that define the clusters are more valuable than the designer graphic displaying the clusters, so going by them south_western is prob between a basque/iberian-type WHG/EEF and a sardinian-type EEF substrate and thus for north_italians (academically based) in specifics it ought to than be more the add sardinian-type EEF substrate rather than the former;

Angela
17-01-18, 20:05
^That is incorrect. In order to interpret Italian genetics you have to take into account known facts of Italian history and pre-history.

There has been no significant inflow of genetic material from Sardinia to mainland Italy since the time of the Nuragic trade contacts, which would have left trace amounts if anything and, moreover, would not have affected most of Northwestern Italy. What gene flow there has been would have been the other way, as is also the case with Corsica, but in both cases is mostly quite recent, and in the case of Sardinia, at least, probably didn't penetrate to the isolated interior plateau which is the source of the "Sardinian" samples in academic sources.

The similarity between Sardinians and Northern Italians is because of shared genetic material from the Neolithic.

Since then Sardinia has been isolated not just in terms of inflow but of outflow. I don't know how to say it more clearly.

The genetic influences on Northwestern Italy are, of course, the Neolithic, the "Ligurians", who were most likely an Indo-European speaking group and were held to inhabit broad areas of Northwestern Italy, Southern France, and into Spain, some Greek input from Massalia and smaller emporia, and some Lombard influence.

For the "Ligures"...

"Traditional accounts suggested that the Ligures represented the northern branch of an ethno-linguistic layer older than, and very different to, the proto-Italic (https://en.wikipedia.org/wiki/Proto-Italic_language) peoples. It was widely believed that that a "Ligurian-Sicanian" culture occupied a wide area of southern Europe,[12] (https://en.wikipedia.org/wiki/Ligures#cite_note-12)stretching from Liguria to Sicily and Iberia. However, while any such area would be broadly similar to that of the paleo-European "Tyrrhenian culture (https://en.wikipedia.org/wiki/Tyrrhenians)" hypothetised by later modern scholars, there are no known links between the Tyrrenians and Ligurians.In the 19th century, the origins of the Ligures drew renewed attention from scholars. Amédée Thierry (https://en.wikipedia.org/wiki/Am%C3%A9d%C3%A9e_Simon_Dominique_Thierry), a French historian, linked them to the Iberians (https://en.wikipedia.org/wiki/Iberians),[13] (https://en.wikipedia.org/wiki/Ligures#cite_note-13) while Karl Müllenhoff (https://en.wikipedia.org/wiki/Karl_M%C3%BCllenhoff), professor of Germanic antiquities at the Universities of Kiel and Berlin, studying the sources of the Ora maritima by Avienus (https://en.wikipedia.org/wiki/Avienus) (a Latin (https://en.wikipedia.org/wiki/Latin) poet who lived in the 4th century AD, but who used as a source for his own work a Phoenician (https://en.wikipedia.org/wiki/Phoenicia) Periplum (https://en.wikipedia.org/wiki/Periplum)of the 6th century BC),[14] (https://en.wikipedia.org/wiki/Ligures#cite_note-14) held that the name 'Ligurians' generically referred to various peoples who lived in Western Europe (https://en.wikipedia.org/wiki/Western_Europe), including the Celts, but thought the "real Ligurians" were a Pre-Indo-European (https://en.wikipedia.org/wiki/Neolithic_Europe) population.[15] (https://en.wikipedia.org/wiki/Ligures#cite_note-15)"

They were both. With the hindsight provided by ancient dna we can now propose this as likely. This happened all over Europe: Incoming Indo-European speakers admixed with prior Neolithic inhabitants.
"Those in favor of an Indo-European origin included Henri d'Arbois de Jubainville (https://en.wikipedia.org/wiki/Henri_d%27Arbois_de_Jubainville), a 19th-century French historian, who argued the Ligurians were the earliest Indo-European speakers of the Western Europe. Jubainville's "Celto-Ligurian hypothesis", as it latter became known, was significantly expanded in the second edition of his initial study. It inspired a body of contemporary philological (https://en.wikipedia.org/w/index.php?title=Philogy&action=edit&redlink=1) research, as well as some archaeological work. The Celto-Ligurian hypothesis became associated with the Funnelbeaker culture (https://en.wikipedia.org/wiki/Funnelbeaker_culture) and "expanded to cover much of Central Europe".[19] (https://en.wikipedia.org/wiki/Ligures#cite_note-19)Julius Pokorny (https://en.wikipedia.org/wiki/Julius_Pokorny) adapted the Celto-Ligurian hypothesis into one linking the Ligures to the Illyrians (https://en.wikipedia.org/wiki/Illyrians), citing an array of similar evidence from Eastern Europe. Under this theory the "Ligures-Illyrians" became associated with the prehistoric Urnfield (https://en.wikipedia.org/wiki/Urnfield_culture) peoples.[20] (https://en.wikipedia.org/wiki/Ligures#cite_note-20)"


"Little is known of the Old Ligurian language (https://en.wikipedia.org/wiki/Ligurian_language_(ancient)). It is generally believed to have been an Indo-European language (https://en.wikipedia.org/wiki/Indo-European_languages) with particularly strong Celtic (https://en.wikipedia.org/wiki/Celtic_languages) affinities, as well as similarities to Italic languages (https://en.wikipedia.org/wiki/Italic_languages). Only some proper names (https://en.wikipedia.org/wiki/Proper_name) have survived, such as the inflectional suffix (https://en.wikipedia.org/wiki/Suffix)-asca or -asco"village".[2] (https://en.wikipedia.org/wiki/Ligures#cite_note-2)"

"The Periplus of Pseudo-Scylax (https://en.wikipedia.org/wiki/Periplus_of_Pseudo-Scylax) describes the Ligyes (Ligures) as living along the Mediterranean coast (https://en.wikipedia.org/wiki/Mediterranean_coast) from Antion (https://en.wikipedia.org/wiki/Antion) (Antibes (https://en.wikipedia.org/wiki/Antibes)) as far as the mouth of the Rhone (https://en.wikipedia.org/wiki/Rhone); then intermingled with the Iberians from the Rhone to Emporion in Spain.[7] (https://en.wikipedia.org/wiki/Ligures#cite_note-7)"

https://en.wikipedia.org/wiki/Ligures

Regio X
17-01-18, 20:14
it only looks like 23andme pulls your north italy down to 13% ...ftdna has you higher.
it also looks like the more west you go from Veneto/friuli/trentino towards turin and Genoa the more higher the North Italian % becomesPossibly SW European also tend to be higher the more west you go in N. Italy. My father is 1/4 MN in ancestry, which maybe help to explain his higher SW.E. % in Gencove.

Pax Augusta
17-01-18, 20:54
The similarity between Sardinians and Northern Italians is because of shared genetic material from the Neolithic.

Exactly, and it was a gene flow from mainland Italy (north and central Italy) to Sardinia and not vice versa.

And modern-day Sardinians have more variation than the usual HGDP sample show.

alexfritz
17-01-18, 21:29
The similarity between Sardinians and Northern Italians is because of shared genetic material from the Neolithic.

exactly that, and never said otherwise, in fact #37 "with a strong affinity towards modern sardinians which is neolithic based" 20-12-17; yes the ligurians were a very large population and it is known archaeologically since the terremare that the po valley could foster/harbor such large populations, which i see as heavily EEF ?parma_beaker pos being the ligurian type? and modern north_it substrate, but actual aDNA of ligurians is still missing;

AdeoF
18-01-18, 04:28
Possibly SW European also tend to be higher the more west you go in N. Italy. My father is 1/4 MN in ancestry, which maybe help to explain his higher SW.E. % in Gencove.
Yep also with Iberians which is shown in gedmatch. Sardinians and other southern Europeans have a good amount of neolithic which also explains that

Angela
01-02-18, 23:24
This is a link to an interview with Joe Pickrell, the scientist behind gencove. It's part of a series that Razib Khan and Spencer Wells are doing at Insitome.

http://insitome.libsyn.com/website/ancestry-deconvoluted

They have a number of them, on the Neolithic, the Indo-Europeans, etc.

Salento
02-02-18, 06:41
This is a link to an interview with Joe Pickrell, the scientist behind gencove. It's part of a series that Razib Khan and Spencer Wells are doing at Insitome.

http://insitome.libsyn.com/website/ancestry-deconvoluted

They have a number of them, on the Neolithic, the Indo-Europeans, etc.

I didn’t realize that You posted this already. Can you Please delete the thread that I started about this Topic. I’m not sure how to do it. Sorry. Link:
https://www.eupedia.com/forum/threads/35351-Insitome-Podcast

italouruguayan
28-03-18, 18:27
My Gencove results:

48% Northern Italy
18% Americas
15% Northern and Central Europe
5% Middle East
4% Ashkenazi Jew
3% South Western Europe
3% Eastern Africa

FTDNA

South East Europe 32%
British Isles 19%
New World 19%
Iberia 13%
East Europe 9%
Middle East 3%
Trace results (1-2%) Ashkenazi , East African

DNA Land

South/Central Europe 47%
North West Europe 20%
Native American 17%
Arab/Egyptian 5%
South Western Europe 4%
East African 3%
Ashkenazi 2%

I am surprised by the extreme variability of the Iberian component (FTDNA 13%, Eurogenes K36 14% Gencove and DNA land 3-4%)