Dodecad Dodecad euro7

Actually, although your numbers definitely have a "Northern Italian" look to them, they are a bit different from the published averages for northern Italians.

Your Northeastern number is very high, for example, especially for someone from Liguria, and yet the Caucasus number is a little high for a northerner as well, and I've never seen a Ligurian or Lombard or Tuscan for that matter score an African number. The only "northern Italian" group that seems to score an African percentage is the OT group. (They get .8 for it as an average.) I just assumed that was a group of mixed southern and northern Italians, but Sile posted that they're from the far northeastern mountainous areas, near Austria. I have no idea what population movement could have been responsible for it, but the neighboring Tyrol also harbors some rather "exotic" clusters for that geographical area. I mention it because I think I recall that you posted that one of your grandparents is from the northeast? Perhaps from that area?

Here are the actual averages from Dodecad "Euro 7" for the northern Italians.

HGDP sample/Bergamo
SW 34.3
NW 31
SE 27.9
NE 4.8
Caucasus 2
Far Asia 0
Africa 0

Dodecad Project North Italians
NW 31.4
SE 28.8
SW 27.9
NE 9.4
Caucasus 2.5
Far Asia 0
Africa 0
I'm not fully ligurian indeed. My maternal grandfather was from Padua, Veneto. One of my maternal great grandmothers was from the emilian appennine near Modena. My paternal grandma was from Parma, Emilia. My part venetian ancestry could explain the higher than average north eastern euro, I suppose.
My "african" 1% is a total mystery to me. Other Dodecad calculators call it "palaeo african" so it should be a very ancient component. I have two theories: maybe it's just some noise (other calculators even found arctic and amerindian...) you know Dodecad is not a scientific project, just amateur. Since other calculators call it "paeleo african", maybe it's a very ancestral component which survived to millennia of admixture instead, only God knows why.
 
Ask him what the terms mean or just get tested by a proper institution instead of an internet Blog;
You are right, unfortunately GEDmatch doesn't support Geno 2.0 kits (so they told me via e-mail), and dr. Mc Donald doesn't test full europeans anymore (he said he should be paid in that case, because of the high workload). I tried other Dodecad calculators with the DIY software and they found totally different results from each other (for example, one even found some traces of "arctic" and "amerindian", which make no sense at all).
EDIT: I was wrong, the african component in the other DIY calculators is called north west african not palaeo african, all of these names are a bit confusing...
 
Arctic and Amerindian actually represent ancient traces of East Asian-like ancestry, it shouldn't be taken too literally. In Southern Europe the figures tend to be pretty irrelevant, yet in Northern Europe the connection seems definitely stronger.

In this test, the Far Asian category is the best proxy regarding these admixtures. Considering your result (0%), that gives an idea of its little impact.
 
You are right, unfortunately GEDmatch doesn't support Geno 2.0 kits (so they told me via e-mail), and dr. Mc Donald doesn't test full europeans anymore (he said he should be paid in that case, because of the high workload). I tried other Dodecad calculators with the DIY software and they found totally different results from each other (for example, one even found some traces of "arctic" and "amerindian", which make no sense at all).
EDIT: I was wrong, the african component in the other DIY calculators is called north west african not palaeo african, all of these names are a bit confusing...


Each Dodecad run was created to explore different aspects of pre-historic population movements, so each has different framing populations and different numbers of clusters etc.That's why you see different "populations" in each run. If you want to get a handle on the different ones, I would suggest you google them at the dodecad blog or Dienekes' blog, or at 23andme. You'll find comments in those places from people who are familiar with the runs, how they work, and what they mean.

Have you run the Globe 13? Given the framing populations, and the clusters, I think it's the most informative for that kind of ancient ancestry in Europeans. This is the spreadsheet with the population averages:
https://docs.google.com/spreadsheet/ccc?key=0ArAJcY18g2GadF9CLUJnTUdSbkVJaDR2UkRtUE9kaUE#gid=2

The results are grouped by source, so you can find Italian results, for example, in the Dodecad section, Hap Map, Li et al etc. The same is true for other groups.

In terms of more recent ancestry, I have to be honest and say that I'm not impressed by the Geno 2.0 analysis, either for uniparental markers or for autosomal analysis, but it should pick up more recent traces. If you really want to play with the data and the DIY, you might consider buying the 23andme service and then running the DIY calculators with that data and then look for correlations.

Just a word of caution if you should test at 23andme and the results should show an up to 1% SSA score. Normally, if "colonial" Americans test there and get a score like that, they would estimate that perhaps a ggggg grandparent was African, and probably a slave. These low level kinds of ancestry results would take them back to the 1700's to the early 1800's, which would be about right.

That doesn't always work with Europeans. As an example, it's not rare for some Sicilians to score anywhere from .3 or .4 up to 1% total SSA or even a little higher at 23andme. Yet, some of these people come from isolated interior villages where any recent such ancestry would be known. (You can't hide anything in places like that, as I'm sure you know.) Plus, it's very widespread over the island. There was no influx of Africans into Sicily within the last two hundred years. (Well, other than very recently.) The last population movement into Sicily which could have carried it all over the island would have been the "Moorish" invasion which covered the period from about 827 to 1061. So, despite their claims, probably to be very conservative, that their analysis goes back only 500 years or so, it seems clear to me that it is picking up SSA from much further back. The fact that it still does show up indicates to me that in more isolated, inbred areas the segments can keep getting passed back and forth. Also, I think the fact that it is retained may mean that it is linked to genes which are providing some sort of selective advantage in a particular environment. I've wondered, for example, whether it might be retained in areas that were affected by malaria until quite recently.

FWIW, from comparing the Dodecad results on something like Globe 13 to those of 23andme, there is definitely a good correlation in terms of these minority ancestries. Of course, that doesn't mean it's always the case, or always going to be the case.
 
You are right, unfortunately GEDmatch doesn't support Geno 2.0 kits (so they told me via e-mail), and dr. Mc Donald doesn't test full europeans anymore (he said he should be paid in that case, because of the high workload). I tried other Dodecad calculators with the DIY software and they found totally different results from each other (for example, one even found some traces of "arctic" and "amerindian", which make no sense at all).
EDIT: I was wrong, the african component in the other DIY calculators is called north west african not palaeo african, all of these names are a bit confusing...

Sounds like the Mozabites i like so much .... ... ...... ...... ; And even if it costs (and you want to have the results) def. go for a proper institution and judge by proper academic studies;

Italians (North Italians / Tuscans / Sardinians / South Italians / Sicilians) are highly diverse from each other (genetically not related to each other DiGaetano et al 2012 due to not inter-mixed with each other Coop and Ralph et al 2013) so if your ancestry is all North Italian than you should cluster with them acc. to K12b North Italy HGDP/Stanford Uni. [11 samples] was 0.7% Mozabite and 0% all other African;
 
This test is useless for anyone who wasn't in the original run that the allele frequencies came from.

That's because most of the clusters are separated by low Fst (genetic) distances and this results in a horrible calculator effect (very similar to PCA projection bias) that makes Italians look Austrian, the English German, and so on.

I can't believe a lot of people don't understand this problem yet, even after being part of the scene for years. Maybe a new hobby is in order, like collecting beer cans or watching grass grow?
 
Dodecad is crap. I tested my autosomal with Family Finder and they found no african, arctic neither amerindian in me (I'm full euro from Northern Italy). They found me 85% european (69% north mediterranean and 16% european northlands), and 16% Anatolia + Caucasus.
No SSA, no amerindian, no arctic, as simple logic could easily suggest...
cut_myorigins.jpg
 

Attachments

  • ethnic makeup marco.jpg
    ethnic makeup marco.jpg
    37.6 KB · Views: 137
Last edited:
This test is useless for anyone who wasn't in the original run that the allele frequencies came from.

That's because most of the clusters are separated by low Fst (genetic) distances and this results in a horrible calculator effect (very similar to PCA projection bias) that makes Italians look Austrian, the English German, and so on.

I can't believe a lot of people don't understand this problem yet, even after being part of the scene for years. Maybe a new hobby is in order, like collecting beer cans or watching grass grow?

someone has to look central alpine european! .............where are the Austrians, swiss, tyrol people in these tests?

We cannot leave these people blank like all the other testing bodies, ie, 23andme, etc
 
Dodecad is crap. I tested my autosomal with Family Finder and they found no african, arctic neither amerindian in me (I'm full euro from Northern Italy). They found me 85% european (69% north mediterranean and 16% european northlands), and 16% Anatolia + Caucasus.
No SSA, no amerindian, no arctic, as simple logic could easily suggest...
View attachment 6476

For myOrigins, you look typical north-italian ( a bit of ancient antolian ), med basin and you should have some german.

is myOrigins correct............doubt it
 
For myOrigins, you look typical north-italian ( a bit of ancient antolian ), med basin and you should have some german.

is myOrigins correct............doubt it
I don't. It's made by professional genetists, not bloggers (no offense folks! :))
 
Northwestern: 31,96
Northeastern:17,23
Southeastern:20,00
Caucasus:5,4
Southwestern:16,77
Far Eastern: 7,98
African:0,58
 
i really wanted to try this calculator, downloading the diy base and then winzip to unzip and then opening the readme and finding i had to download something called 'r' gave me an ulcer. will we ever see any of these calculators on gedmatch?
 
But, i think its very common in northern Europe, such as northern Sweden, to have both east asian, native, and some south European as well. Usually says Iberian.
 
Greek (father from Thessaly, mother from Central Greece):

Admix Results (sorted):


# Population Percent
1 Atlantic_Baltic 46.01
2 Southern 28.52
3 West_Asian 25.11
4 Siberian 0.37


Single Population Sharing:


# Population (source) Distance
1 Tuscan (HGDP) 5.82
2 O_Italian (Dodecad) 5.95
3 Greek (Dodecad) 6.81
4 C_Italian (Dodecad) 7.16
5 TSI30 (Metspalu) 7.22
6 Bulgarian (Dodecad) 9.06
7 Bulgarians (Yunusbayev) 9.97
8 Romanians (Behar) 11.05
9 N_Italian (Dodecad) 12.37
10 S_Italian_Sicilian (Dodecad) 12.96
11 North_Italian (HGDP) 13.39
12 Sicilian (Dodecad) 14.08
13 Ashkenazy_Jews (Behar) 15.16
14 Ashkenazi (Dodecad) 15.37
15 Baleares (1000Genomes) 19.86
16 Murcia (1000Genomes) 20.45
17 Canarias (1000Genomes) 20.72
18 Andalucia (1000Genomes) 20.83
19 Extremadura (1000Genomes) 20.96
20 Galicia (1000Genomes) 21.26
 
genesis.gedmatch.com_fcgi-bin_admixProp.cgi%2B%25281%2529.png
 
Last edited:

This thread has been viewed 61980 times.

Back
Top