search this blog

Monday, May 11, 2015

4mix: four-way mixture modeling in R


Thanks to Eurogenes project member DESEUK1. A zip file with the R script, instructions and a couple of data sheets is available here.

So let's model Poles as a bunch of ancient genomes from Central and Eastern Europe using output from my K8 analysis.

Copy & Paste: source('4mix.r')

Hit ENTER

Copy & Paste: getMix('K8avg.csv', 'target.txt', 'HungaryGamba_EN', 'HungaryGamba_HG', 'Karelia_HG', 'Corded_Ware_LN')

Hit ENTER

After a few seconds you should see the results...

Target = 19% HungaryGamba_EN + 14% HungaryGamba_HG + 2% Karelia_HG + 65% Corded_Ware_LN @ D = 0.0062






Obviously the script can use ancestry proportions and/or population averages from any test, provided they're formatted properly. The accuracy of the modeling will depend on the quality of the input.

Update 19/05/2015: A new version of the 4mix script that can run multiple targets is available here, courtesy of Open Genomes.

204 comments:

«Oldest   ‹Older   201 – 204 of 204
Krefter said...

@Davidski,
"However, keep in mind that southern Europe, especially southeastern Europe, has seen gene flow from the post-Neolithic and even post-Islamic expansion Near East"

Can you give a few pre-Islamic examples of Near Eastern admixture in South Europe?

We shouldn't be surprised that Greeks are intermediate between mainland Europeans and West Asians, considering where it is geographically. We could say Greece instead has post-Neolithic European admixture. We don't have ancient Greek genomes.

Just because Greeks and South Italians cluster with each other, doesn't mean they got their West Asian-affinity from the same source. European Jews cluster with Greeks, etc. but obviously for differnt reasons.

Davidski said...

Sure, how about the Phoenicians?

http://en.wikipedia.org/wiki/Phoenicia

Matt said...

@ Capra, late to reply, a little off topic, but interesting other stats from that list (http://textuploader.com/ps1c)

Chimp Ust_Ishim LBK_EN Australian -0.0036 -0.615 352857
Chimp Ust_Ishim LBK_EN Bougainville 0.0012 0.228 352858

So whatever admixture / divergence Oceanians have which take them away from ENA seems comparable in effect to what LBK in some ways...

Another odd pattern is the way that

Chimp Ust_Ishim LBK_EN Punjabi 0.0108 3.379 352858

Chimp Ust_Ishim LBK_EN GujuratiB 0.0074 3.379 352858

(compared to the Yamnaya stat you've put down)

which is again pretty weird if only the absolute level of Near East / ENF components is all that matters and not the balance of ENA vs EHG vs WHG for affinity to Ust Ishim (and if the predictions are that Punjabi / Gujurati B should have more Near East / ENF than Yamnaya according to these ADMIXTURE).

But then the WHG / EHG groups do seem less close to Ust Ishim than the East Asian and Native Americans via the same stat, anyway (see above text).

So perhaps UI is not such a great tool for discovering Basal Eurasian level, as like using ENA, it is still affected, albeit at a much lower level, by recent ENA admixture / something else, not just level of Basal Eurasian.

Quick visualisations of a few stats - http://i.imgur.com/2wV0gfu.png / http://i.imgur.com/tQkiWvB.png / http://i.imgur.com/oeh1hx9.png

Simon_W said...

I think the K8 is better than the K9, it makes the more sensible results. My K8 results make complete sense in the light of my latest theories. One of my best K8 approximations:

26% West_Scottish + 47% HungaryGamba_BA + 2% Belorussian + 25% Tuscan @ D = 0.007

Thus, my roughly 50% South German and German Swiss ancestry is entirely approximated by HungaryGamba_BA, which apparently is a good fit for central Europeans. My East Prussian quarter is approximated by the Western Scottish (the old Saxons were strongly Northwest European, see Hinxton2) with some low Belorussian-like admixture. My Italian grandfather is approximated by the Tuscans.

When I use the Corded Ware instead of the Western Scottish I get an even better approximation, though slightly different:

10% Corded_Ware_LN + 1% Belorussian + 67% HungaryGamba_BA + 22% Tuscan @ D = 0.0069

Now HungaryGamba_BA has increased and partly corresponds also to ancestry contributed by my non-central European grandparents. Makes sense, because Bronze Age movements from Hungary or more generally from central Europe to northern, northwestern and southern Europe were non-negligible. HungaryGamba_BA in Italy is old Italic influence and similar people admixed into the Germanics. What differentiates Northwest Europeans like the old Saxons from modern central Europeans is an extra batch of Corded Ware ancestry. And what differentiates Italians from central Europeans is a Cypriot-like influence, as can be seen in this series of approximations:

9% Corded_Ware_LN + 2% Belorussian + 78% HungaryGamba_BA + 11% Cypriot @ D = 0.0071

10% Corded_Ware_LN + 0% Belorussian + 76% HungaryGamba_BA + 14% South_Italian @ D = 0.0071

9% Corded_Ware_LN + 0% Belorussian + 73% HungaryGamba_BA + 18% Greek @ D = 0.007

9% Corded_Ware_LN + 1% Belorussian + 73% HungaryGamba_BA + 17% Italian_Abruzzo @ D = 0.0069

10% Corded_Ware_LN + 1% Belorussian + 67% HungaryGamba_BA + 22% Tuscan @ D = 0.0069

The quality of the fit only changes marginally, but the portion of the admixture decreases from Tuscan continuously to Cypriot. Which means that this type of admixture is most concentrated and pure in Cypriots. I guess this is connected with tales of "Pelasgians" and the spread of J2 + old R1b + ANE-admixed West Asians.

«Oldest ‹Older   201 – 204 of 204   Newer› Newest»