I don't think you should give up on trying the 3-way. Or have you tried the reference populations that I shared when I posted mine?Quote:
Originally Posted by Croatian Legend
Printable View
I don't think you should give up on trying the 3-way. Or have you tried the reference populations that I shared when I posted mine?Quote:
Originally Posted by Croatian Legend
yes i tried it also it is also completly wrong. See that is my point you cannot use a your refernce with my genetic background. Also those 3 simple models are also not always good because looking at yamnaya it is basically 50/50 what if you have elevated CHG like me from an old dinaric balkan source. This is why you have to seperate the chg and ehg to see the actually breakdown. That is why the p value improves.Quote:
Originally Posted by Gannicus
Just to make that clear that ido have steppe in my but i also have extra chg which is explained by scientific papers that go into detail about the population in the dinaric plain ecpecially dinaric croats! Vahaduo inflates eef because it fails to seperate chg and other stuff as explained before.Quote:
Originally Posted by Gannicus
You're right that some of my references won't work for you but sometimes it can with enough tweaking. Just keep tinkering!Quote:
Originally Posted by Croatian Legend
Well i achieved the a perfect almost perfect model for me 4 way. That comes in line with scientifc claim and research. Sure i will try 3 way but 3 way is not the way to go in my case escpecially with elevated CHG .Quote:
Originally Posted by Gannicus
What refrences and sources do you all recommend for people in the balkans, just got the qpAdm in illustrativedna but dont really know what to use for most realistic results
If you have seen any of AndreiDNA's videos, you can see in some of the qpAdm data the references. You can start with what he uses in some of the models he has used and then make your own adjustments.Quote:
Originally Posted by MandM
I've also had him model me before, and to start I used the references he used:
Attachment 142481
Here are some models of myself with Sintashta Culture. It's widely accepted that the Sintashta people are the inventors of the chariot.
Attachment 142483
Attachment 142484
Attachment 142485
Attachment 142486
Runs where the sources with Sintashta failed:
Attachment 142487
Attachment 142488
References used:
Attachment 142489
Attachment 142490
Now on to the Battle Axe culture!
BAC + Sicily EBA
Attachment 142504
BAC + Ireland Megalithic
Attachment 142505
BAC + England N Megalithic
Attachment 142506
BAC + Greece EBA Manika
Attachment 142507
BAC + Levant N
Attachment 142508
References used:
Attachment 142509
Attachment 142510
Attachment 142511 I’ve got more corded ware models of myself that I’ll share later tonight or the coming days. Later I’m going to experiment with the specific EEF archaeological cultures like Cardial, globular Amphora, Remedello, Vinca etc + Yamnaya and cordedware.
Recommend this model for South Slavs
https://i.postimg.cc/3NbMsG9P/temp-Imageb-Coo-JS.avif
Extended Distance list, looking at the West Asian Bulgarian as closest makes sense
Code:0 Bulgarian.HO 0.0003
1 Croatian.HO 0.0009
2 Ukrainian.HO 0.0017
3 Hungarian.HO 0.0021
4 Italian_North.HO 0.0030
5 Belarusian.HO 0.0031
6 French.DG 0.0036
7 Spanish.DG 0.0040
8 Sicilian.HO 0.0040
9 Greek.HO 0.0040
10 Norwegian.HO 0.0049
11 English.DG 0.0063
12 Lithuanian.HO 0.0064
13 Russian.DG 0.0066
14 Armenian.DG 0.0069
15 Lezgin.DG 0.0071
16 Cypriot.HO 0.0072
17 Kumyk.HO 0.0081
18 Karelian.HO 0.0086
19 Jew_Ashkenazi.HO 0.0092
20 Orcadian.DG 0.0104
21 Iranian.HO 0.0119
22 Abkhasian.DG 0.0120
23 Basque.DG 0.0124
24 Sardinian.DG 0.0131
25 Egyptian.HO 0.0168
26 Palestinian.DG 0.0169
27 Berber_WGA.HO 0.0288
28 Saami.DG 0.0304
29 Uyghur.DG 0.0415
Estonia CordedWare + Spain EBA
Attachment 142530
Estonia CordedWare + Minoans
Attachment 142531
Estonia CordedWare + Germany LBK N (EEF)
Attachment 142532
Estonia CordedWare + Levant N (Jordan PPNB)
Attachment 142533
Estonia CordedWare + Israel Chalcolithic
Attachment 142534
I modeled myself with a couple Bellbeakers: Attachment 142536Attachment 142537Attachment 142538 One thing to note is that Italy Bronze Age sample I think is from the south around Puglia and has a decent amount of steppe ancestry.
Would you mind sharing the reference populations you used for these?Quote:
Originally Posted by Gannicus
Attachment 142542Attachment 142543Attachment 142544Quote:
Originally Posted by Xavier727
[QUOTE=Gannicus;8276553]Attachment 142542Attachment 142543Attachment 142544[/QUONice oneTE]
Nice, thanks man
[QUOTE=Xavier727;8276579]You're welcome.Quote:
Originally Posted by Gannicus
Did some models of myself with Steppe + EEF Cultures + sometimes WHG
Steppe + Cardial Ware + WHG
Attachment 142546
Steppe + Remedello Culture
Attachment 142547
Steppe + Globular Amphora Culture
Attachment 142548
Steppe + Vinca Culture + WHG
Attachment 142549
Steppe + Funnel Beaker Culture
Attachment 142550
References:
Attachment 142551
Attachment 142552
Was able to pull out Natufian in this run:
Steppe + ANF + Natufian + WHG:
Attachment 142578
References:
Attachment 142579
Attachment 142580
With the same references I did the same model but with Boncuklu PPN instead of Barcin:
Attachment 142581
Quote:
Originally Posted by Gannicus
Have you managed to come up with any decent medieval models yet? I’ve been trying but haven’t had much luck.
Not yet. The only one that comes close that has worked is this one: Attachment 142589 These two are right on the edge of being early medieval. And I need to experiment more with references. The others I have tried so far have failed.Quote:
Originally Posted by Xavier727
Quote:
Originally Posted by Gannicus
Yeah, this is about the best I’ve managed so far, but it still needs some tweaking.
Attachment 142590
Attachment 142591
Attachment 142592
Attachment 142593
I just did this one and again with high SEs. Trying some other combinations with Ireland Kilteasheen before I start tweaking the references.Quote:
Originally Posted by Xavier727
Attachment 142594
Current references for this:
Attachment 142595
Attachment 142596
With the same reference list as in post #144:
Attachment 142597
This one is with Spain BA and Sicily EBA added to the references:
Attachment 142598
Removed Jordan PPNB and added Denmark LN BA VK 214:
Attachment 142599
I don't think this is a good model guys
Attachment 142604
I have a couple clean qpAdm models of myself with Medieval groups now:
Attachment 142608
Attachment 142609
The pooled Italy Medieval samples are Southern Italians.
references:
Attachment 142610
Attachment 142611
Have you done an Iron age model?Quote:
Originally Posted by Gannicus
A few. And one thing I find interesting is that IA England samples don't work too well for me. I can get some combinations to work. There is correlation between G25 and qpAdm I have noticed but there are some differences.Quote:
Originally Posted by drb234
Here are some IA models I have done:
Attachment 142615
Attachment 142616
Attachment 142617
Attachment 142618
Attachment 142619
I've modeled myself with Bronze Age populations too. My genome seems to pair well with Ireland EBA Rathlin 2 so I have used that individual in a lot of runs.Quote:
Originally Posted by drb234
Attachment 142629
Attachment 142630
Attachment 142631
Attachment 142632
Attachment 142634
Let me know if you have any populations you would like me to try.
This model is where I was experimenting with Serednii Stih/Sredny Stog.
Attachment 142635
Attachment 142636
So here i managed to lower the standard error with both AHG and CHG whilst retaing excellent p-value. And i have found out that i have definitly an extra CHG component from the Caucasus which also confirms the findings by professional papers. The AHG went up but stayed around the 35% mark even after trying various runs.
Well it is true Croats that have roots from the Lika/Dinar area have elevated CHG and low EEF. The catch though is that the CHG component is elevated!
The only thing i am still trying to figure out why i am better modelt with the precursor to EEF. (Researching that already and reading some stuff om that.)
That CHG you have there is before Kotias, which in the mesolithic. That late UP sample could be soaking up some Iran Neolithic. The CHG that is typically used in these models is Georgia_Kotias_Mesolithic.Quote:
Originally Posted by Croatian Legend
Nope not at all! I tried Kotias and the P value goes drastically down. I seem to have a more pure CHG variant and Basal reacher version. Kotias is already mixed with a litte EHG. To get a more cleaner split satsublia is better to get the exact CHG out especially someone like me that has a more Caucasian/Steppe related input.Quote:
Originally Posted by Gannicus
You seem to think in terms of G25/ Vahaduo ways, there is no one fit all calculation. You have to be more logical and flexible especially in a comlpex region like the Dinaric area.
I am going to repeat once more i tried those standard setups they FAIL with me!
You see this is what alot amateurs here donīt understand instead spewing BS theories and using Vahaduo as proof is laughable.
There are many scientific papers out there for various European groups and the genetic breakdown is a litte more complex then people here are aware.
No attack on you just a heads up.
No, what I'm saying is that Kotias is used in population genetics studies. When you hear discussions about "CHG" admix it's usually Kotias. What does the p-value go down to when you use Kotias?Quote:
Originally Posted by Croatian Legend
G25 I don't think deserves all the hate it gets. It is based on PCA which is what academics use. It for sure takes a back seat to academic tools like qpAdm. I've observed qpAdm behaving a lot like G25. I see correlation between the two with my own results. It's not an exact 1:1, but a similar pattern emerges.
I noticed that too. Most similarly to qpadm, I am modeled by IllustrativeDNA and G25 Davidski, there is always some margin of error, but still..Quote:
Originally Posted by Gannicus
I think its fine. Should you use it to attempt to refute the findings of a genetics paper? Absolutely not. Other than that I think it can reveal deep ancestry. With some G25 lists, such as the ones with steppe, ANF, WHG, the results are pretty close.Quote:
Originally Posted by Fistora
Quote:
Originally Posted by Gannicus
Sorry but you have no idea or rather you don't actually read scientfic papers because you keep repeating the same nonsense as most here.
Satsurblia is definitly used and also swapped in profesional testing. Satsurblia indicates a pure CHG line and it is used to better seperater CHG from EHG. Also same goes with AHG and EEF: AHG hast an eastern shift and lacks the natufian component that EEF has. It is also used in population/individual testing to see which fits better.
And no genetic scientist uses G25 for professinal work , since you mentioned that it is based on PCA. You do realise that they have there own ways or software to do the PCA. Which is different then G25 from Davidski.
Or rather let me be more clear Davdskis coords are a joke for ancestral calculation no serous scientist uses that. Not to even talk about his methods how he creates the coords. But i won't go into that now.
G25 is only good for general clustering but not to break down your ancestral components.
Nope i have recreated his results with QPADM and i get very bad P-values and some even failed. He has a definitv biased to frame Poles as some Steppeüber population. Even though Qpadm results show that you guys are around 40 to 45 even 50% EEF so pretty normal central European.Quote:
Originally Posted by Fistora
As matter of fact when i recreated his results it was barely over a good 0,05 value and the standard error was horrible.
By the way this is no personal attack on you just talking here.
And again looking at all other scientific work not just for Balkans also other regions of Europea you would see that things are a little more complex. I just find it funny that you take G25 from davidski as some proof even though all of the professionals actually warn from using Amateur tools like Davidski because it creates bias and a wrong picture.