I am working on some qpGraphs to figure out the different Iran farmer like ancestries in Ganj Dareh, CHG, TepeAbdulHosein, Hotu cave, Wezmeh cave, SC Asia (Sarazm) and India etc. Its taking longer than I expected because of many populations adding complexity.
So I thought i would use the graphs already made to shed some light on some Eurasian samples. When I publish qpGraphs please keep some things in mind while analyzing.
a. I look to reduce the worst f4 stats. The worst fstat will be on top of the graph, aim is to keep it under 3.
b. My passing model is probably no going to be the only passing model.
c. Shaded rectangle boxes are actual samples used as input fro qpGraph, other oval nodes are hypothetical populations for which we may or may not have matching actual samples in the aDna database. Even if they are, I definitely haven't used them as input for this graph.
d. qpGraph changes as new samples are added. For eg, Tianyuan being a source for ANE in below graph might not hold as true if more east eurasian populations are added, instead another eastern source might be selected.
e. Sample notation is country, sample name, date.
f. I'm still figuring out resolution issues for graphs, although clicking the graph and seeing the high res image on mobile should offer best view of the image.
g. Solid lines are drift edges, each node can only have 2 drift edges coming from it as per qpGraph design. The numbers on drift edges is drift in drift units. Dotted lines are admixture edges with admixture % noted on top of the dotted line.
Some observations based on graph
1. Romania Oase has 6-7% excess neanderthal ancestry.
2. West Eurasians Kostenki, Sunghir and Villabruna might have upto 1% Neanderthal ancestry.
3. Kostenki & Sunghir are on same West Eurasian branch.
4. ANE is 75-82% West eurasian and 25-18% east eurasian (eastasianinANE node). The 18% east eurasian is from both tianyuan like and onge like sources (will change as more eastern ancients are added)
5. The same East eurasian mix needed for ANE/MA1 Malta boy is a good fit for the excess east eurasian in Tarim_EMBA. There is excess 20-25% east eurasian in Tarim_EMBA over MA1.
6. Tarim_EMBA also has around 3% IranN like ancestry and 11% Villabruna related ancestry. If someone can explain to me the Villabruna part I shall be grateful.
7. Yana is 30% east Eurasian, 35-36% ANE, 35% west eurasian (Sunghir).
8. IranN requires basal eurasian ancestry (nodes NonAfrican or MainEurasian), model fails with other sources.
Anything else that you find interesting, let me know.
Thoughts?
ReplyDeleteFrom the above model looks like Pre villabruna had a more easterly spread than its descendent.
ReplyDeleteCongratulations for starting your own blog. A good decision !
ReplyDeleteAbout this particular qgraph, I am not sure why Villabruna manages to remain a pure unadmixed West Eurasian population. I don’t think that it is a representation of reality as must have been happened.
From what I can recall from a few years back when the European Ice Age paper came out, I think in 2016, Villabruna had a basal R1b lineage but it was perhaps downstream from the Tarim marker. Moreover, the group represented by Villabruna,i.e. Mesolithic Europeans were modelled as having had an admixture from the Near East from a population that had some ANE affinity. I am not sure if I recall all the things accurately here so I would recommend looking back at the paper again. But I believe it is possible that the migration that brought CHG and Iran N into the Caucasus and Zagros respectively may have also contributed to Villabruna. The origin of that source could be from further east than Iran.
Congratulations for starting your own blog. A good decision !
ReplyDeleteAbout this particular qgraph, I am not sure why Villabruna manages to remain a pure unadmixed West Eurasian population. I don’t think that it is a representation of reality as must have been happened.
From what I can recall from a few years back when the European Ice Age paper came out, I think in 2016, Villabruna had a basal R1b lineage but it was perhaps downstream from the Tarim marker. Moreover, the group represented by Villabruna,i.e. Mesolithic Europeans were modelled as having had an admixture from the Near East from a population that had some ANE affinity. I am not sure if I recall all the things accurately here so I would recommend looking back at the paper again. But I believe it is possible that the migration that brought CHG and Iran N into the Caucasus and Zagros respectively may have also contributed to Villabruna. The origin of that source could be from further east than Iran.
Vasishta come to my discord, there is some interesting discussions there. We are talking about IE genetics.
ReplyDeleteIts easy to join just set up an account if you dont have one.
https://discord.gg/yQ5ueUef
Everyone is welcome.
@jaydeep.
ReplyDeleteThanks.
"About this particular qgraph, I am not sure why Villabruna manages to remain a pure unadmixed West Eurasian population. I don’t think that it is a representation of reality as must have been happened."
Yea the qpgraph is constrained by the samples I input in. This does not mean that there is pure villabruna in tarim adna but some later population which shares alleles with villabruna. As we know villabruna is the WHG (iron gates, west European) cluster.
If I add more west European Neolithic populations, the villabruna affinity might go away and be replaced by some other population.
ReplyDeleteQpAdm can also tell us more, however I am really not motivated right now to work on that Question, busy with something else on qpGraph.
ReplyDelete@vasishta
ReplyDeletewhich is the best source for the west eurasian contribution to ANE (82%) Kostenki, Sungir, Goyet, Pestera Muierii....?
@old europe
ReplyDeleteas far as my analysis goes, i didnt test with sunghir, goyet or pestera added.
WE source for ANE is basal to kostenki, villabruna & vestonice.