Its finally time for a new update!
Whats new?
- Included reference populations from all over europe from Spain in the west to Cypriots in the south to Caucasus in the south-east and Chuvashes in the far east.
- ADMIXTURE run from K2 to K5
- MDS plots (earlier wrongly called PCA plots) up to 9 dimensions
- DIST plots similar to what i seen in 23andme. People typically match each other within Europe between 0.78 to 0.79. The differences is in these plots emphasised. There are 4 versions each sortert from the different nationalities Finns, Swedes, Norwegians and Saami.
Whats the result?
This is my suggested interpretation. For some of you these is probably much old news.
ADMIXTURE:
- At K2 or assuming two populations Europe is split into a southern (RED) and northern (BLUE) part with the Sardinians, Cypriots and the Caucasus populations cleary at the southern extreme while the Finns, Saami and the Chuvashes at the northern extreme. North in this context appears to point north-east geographically.
- At K3 or assuming three populations Europe is splitt into 1) south-east Europe populations (BLUE) the the Georgians at the extreme 2) south-west Europe populations (GREEN) with the Sardinians at the extreme 3) north/north-east Europe populations (RED) with the Saami at the extreme.
- At K4 or assuming four populations Europe is splitt into 1) Lithuanians at the extreme (BLUE) 2) Chuvasshes at the extreme (PINK) 3) Sardinians at the extreme (RED) 4) Georgians at the extreme (GREEN).
- At K5 or assuming four populations Europe splitt into 1) Basque at the extreme (RED) 2) Chuvashes at the extreme (PINK) 3) Lithuanians at the extreme (CYAN) 4) Sardinians at the extreme (BLUE) 5) Georgians at the extreme (GREEN).
COMMENT: At the highest K=5 the ADMIXTURE result *alone* seem to suggest that Norwegians, Swedes, Finna and Saami main ancestry is "Lithuanian" but that Finns and especially the Saami have a considerable "Chuvash" influences.
MDS-PLOTS:
- D1-2: The Norwegian and Swedish common clusters together with Belorussians and Lithuanians appear to bridge the space between Orcadians/French and Finns and Volgoda-Russians. The Saami appears to go further past the latter cluster outside towards the Chuvash. The Hungarians appear to neighbour the swedish/norwegian cluster to the upper-right.
- D1-3: The same story as above, but here the Chuvash share the far lower part of the plot with the Sardinians even the distances are large. Also here the Saami appear to pull from the Russians/Finns toward the Chuvashes. In the earlier plot the Chuvashes shared the higher part of the plot with Cacausus populations even the distances where large. The Lithuanians and Belorussians appear to seperate further lower away from the Finns and the Russians.
- D1-4: The same as above.
- D1-5: The Finns appears to have seperatet from the Russian and appears to pull toward the Saami who are alone at the extreme lower-left of the plot. Else much the same as above.
- D1-6: The Finns cluster with the Chuvashes with the Norwegian/Swedish cluster to the right. Russians and Lithuanians below. Saami alone at the upper-left extreme.
- D1-7: The Finns and the Saami appears to cluster alone to the lower-left. Norwegian-Swedish cluster partly with Belorussians/Orcadians.
- D1-8: Mostly the same as above.
- D1-9: Mostly the same as above.
COMMENT: The plots appears to show a more complex picture of the ancestry and relationship than the ADMIXTURE plots. In the lower dimensions the Saami appears to pull toward the Chuvashes but seem to seperate clearly often togheter alone with Finns in higher dimensions. Norwegians and Swedes appears mostly as expected to bridge between the more western populations with the more eastern.
DIST GRAPH:
The graph shows the average distance between populations seen from the four nationalities Norwegians, Swedish, Finns and Saami.
- Norwegians are more similar too Swedes, Belorussians, Orcadians and Lithuanians than themself (!). (Yes I have checked for errors).
- Swedes closest relatives are the Norwegians, Lithuanians and Belorussians. The most distant the Chuvashes, Cypriots and the Georgians.
- Finns closest relatives are the Lithuanians, Swedes and the Belorussians. The most distant are the Cypriots, Adigey and the Chuvashes.
- The Saamis clostest relatives are the Finns, Swedes and the Belorussians. The least distant the Cypriots, Armenians and the Georgians.
- There is very high correlation in the DIST plots between Norwegians and Swedes at 0.9842, then the Finns-Swedes at 0.9387, then Finns-Swedes at 0.9099, then Saami-Finns at 0.725, then Saami-Swedes at 0.511 and finally Saami-Norwegians at 0.4531.
- Saami appears to have consistently much lower genetic affinity with all other populations than the other three. The exception is to the Chuvashes where all appear to have somewhat similar genetic distance.
- Saami affiliation with continental and especially southern European populations are much smaller than the others.
Whats next?
- Get more samples
- Extend reference samples to include Euroasia
- Identify other analys methods
This concludes this round. There are no local MDS plot this time.
Anders
Hi, your PCA (which would be coincident with your K=3) has called my attention, first as mere something I had to study in some detail (the color and scale made it difficult to understand it on first sight) but then, when I did understand it, it comes out as something I have argued for since years ago: a (basically) triangular shape of the genetics of Europe - which is coincident with a basically triangular shape of the continent as well.
SvarSlettThe three vortices are the Volga, the Basque Country and West Asia (incl. North Caucasus).
It is interesting however to see in this PCA that there is a gap between West Asia & Caucasus and Europeans, even those that, like Romanians or Italians, show a clear tendency towards the Eastern vortex. It rings obvious that this gap is to be filled by Turks and Greeks (not plotted) but still it is quite notable when, by comparison, French, Hungarians and peoples of the Baltic area are comparatively so close (in spite of a similar geographic distance).
This is probably because these three populations should have a single common origin majorly in Late UP/Epipaleolithic Europe, which continuously shows up as a shared "North European" indistinct component, from the Atlantic to parts of Russia, which is quite stubborn and hard to split. This is the same as the component you call "Lithuanian" elsewhere (its true origin is probably at the Rhine-Danube Paleolithic province, from Belgium to Hungary, but whatever).
However you also managed to spot, thanks to the Chuvash sample, a further NE European (maybe Siberian or Central Asian ultimately) component. This is interesting because it further stretches the Baltic/Fenoscandian group in that direction and seems to say something about East European peculiarities. Hmmm...