View Full Version : Average samples in Gedmatch
Lemminkäinen
04-23-2020, 04:16 PM
I made following average samples and you can now test how reliable those ancestry tests are. It looks like the Eurogenes K13 is one of the best Gedmatch analyses for Finns and Balts. I will add more averages in future, Central and South Europeans. Samples are done by selecting random alleles per each locus and the quality depends on the sample number.
FU3882944 Estonian-avg
FW2172699 SWFinnish-avg
LJ9046036 Udmurd-avg
CH1944711 Karelian-avg
WQ7145667 EastFinnish-avg
SQ9458948 Saami-avg
QT4281058 RusPskov-avg
CA7050445 Latvian-avg
I personally use Dodecad
Udmurt
Population
Gedrosia 12.64 Pct
Siberian 22.46 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 9.73 Pct
North_European 44.19 Pct
South_Asian -
East_African -
Southwest_Asian -
East_Asian 4.77 Pct
Caucasus 6.21 Pct
Sub_Saharan -
Pskov
Population
Gedrosia 2.22 Pct
Siberian 2.33 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 18.88 Pct
North_European 66.15 Pct
South_Asian 1.46 Pct
East_African 0.09 Pct
Southwest_Asian 0.30 Pct
East_Asian -
Caucasus 8.55 Pct
Sub_Saharan -
66% North_European is way too high for the vast majority of Russians, so that's hardly representative. Most of us are 55-60%. Some are 61-62% or so but 66% is definitely super high for an average. I've seen a couple of hundreds of samples, so I can speak confidently about it.
Lemminkäinen
04-23-2020, 06:40 PM
I personally use Dodecad
Udmurt
Population
Gedrosia 12.64 Pct
Siberian 22.46 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 9.73 Pct
North_European 44.19 Pct
South_Asian -
East_African -
Southwest_Asian -
East_Asian 4.77 Pct
Caucasus 6.21 Pct
Sub_Saharan -
Pskov
Population
Gedrosia 2.22 Pct
Siberian 2.33 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 18.88 Pct
North_European 66.15 Pct
South_Asian 1.46 Pct
East_African 0.09 Pct
Southwest_Asian 0.30 Pct
East_Asian -
Caucasus 8.55 Pct
Sub_Saharan -
66% North_European is way too high for the vast majority of Russians, so that's hardly representative. Most of us are 55-60%. Some are 61-62% or so but 66% is definitely super high for an average. I've seen a couple of hundreds of samples, so I can speak confidently about it.
As I wrote, the sample quality is based on sample number and Pskov is reptesented by 2 samples. The Finns had 24 samples. But you understood me and my test samples wrong. Look at Oracles; the calculator works fine if the Russian average sample gives Russian as the closest Oracle result, no matter what are those proportions you are looking for. If the Finnish sverage gives Finnish Oracle etc.
Lemminkäinen
04-23-2020, 06:50 PM
Dodecad gives for Pskov following Oracle results. It looks like according Dodecad the Pskov area represents "mixed Slavs" and Russians. Sounds like Russians. Show me your Russians.
1 Mixed_Slav (Dodecad) 5.58
2 Russian (Dodecad) 5.64
3 Polish (Dodecad) 5.7
4 Belorussian (Behar) 6.24
5 Russian_B (Behar) 7.81
6 Russian (HGDP) 8.52
7 Ukranians (Yunusbayev) 8.98
8 Lithuanian (Dodecad) 9.15
9 Mordovians (Yunusbayev) 9.5
10 Lithuanians (Behar) 11.76
Eurogenes K13 gives this. I have also a RussRz average and will look tomorrow if it is more Russian. Anyway, I know that historically Pskov is very close Latvians. Eurogenes has no Latvian samples, so Lithuanian sounds right.
1 Lithuanian 3.44
2 Estonian_Polish 6.79
3 Belorussian 7.22
4 Russian_Smolensk 7.4
Dodecad gives for Pskov following Oracle results. It looks like according Dodecad the Pskov area represents "mixed Slavs" and Russians. Sounds like Russians. Show me your Russians.
1 Mixed_Slav (Dodecad) 5.58
2 Russian (Dodecad) 5.64
3 Polish (Dodecad) 5.7
4 Belorussian (Behar) 6.24
5 Russian_B (Behar) 7.81
6 Russian (HGDP) 8.52
7 Ukranians (Yunusbayev) 8.98
8 Lithuanian (Dodecad) 9.15
9 Mordovians (Yunusbayev) 9.5
10 Lithuanians (Behar) 11.76
Eurogenes K13 gives this. I have also a RussRz average and will look tomorrow if it is more Russian. Anyway, I know that historically Pskov is very close Latvians. Eurogenes has no Latvian samples, so Lithuanian sounds right.
1 Lithuanian 3.44
2 Estonian_Polish 6.79
3 Belorussian 7.22
4 Russian_Smolensk 7.4
There are academic Russians on Vahaduo K12b. And there is also a Latvian average on Vahaduo Euro K13.
You are a bit late for the party, sir ;)
Lemminkäinen
04-23-2020, 09:26 PM
There are academic Russians on Vahaduo K12b. And there is also a Latvian average on Vahaduo Euro K13.
You are a bit late for the party, sir ;)
Oh, did it hurt so much to understand that Pskov is not genetically Russian? They were original dwellers. I am not late sir. I am making my own samples and publish them. Will continue with English provinces etc. You have still pretty much to learn.
Lucas
04-23-2020, 10:17 PM
Gedmatch could remove those samples if you upload too much artificial genomes. They were already attacked for letting such possibility.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6946563/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6992384/
Oh, did it hurt so much to understand that Pskov is not genetically Russian? They were original dwellers. I am not late sir. I am making my own samples and publish them. Will continue with English provinces etc. You have still pretty much to learn.
You got me wrong. I just told you there are updated spreadsheets if you want to check them.
Of course it's genetically Russian, one of the oldest Russian cities. A sample size of 2 is way too small to make any conclusions. But I do know that Pskov is heavily Baltic, no news for me.
Ibericus
04-24-2020, 07:34 AM
I personally use Dodecad
Udmurt
Population
Gedrosia 12.64 Pct
Siberian 22.46 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 9.73 Pct
North_European 44.19 Pct
South_Asian -
East_African -
Southwest_Asian -
East_Asian 4.77 Pct
Caucasus 6.21 Pct
Sub_Saharan -
Pskov
Population
Gedrosia 2.22 Pct
Siberian 2.33 Pct
Northwest_African -
Southeast_Asian -
Atlantic_Med 18.88 Pct
North_European 66.15 Pct
South_Asian 1.46 Pct
East_African 0.09 Pct
Southwest_Asian 0.30 Pct
East_Asian -
Caucasus 8.55 Pct
Sub_Saharan -
66% North_European is way too high for the vast majority of Russians, so that's hardly representative. Most of us are 55-60%. Some are 61-62% or so but 66% is definitely super high for an average. I've seen a couple of hundreds of samples, so I can speak confidently about it.
the north-euro component in dodecad is based on Lithuanians, so its normal that russians get high.
Lemminkäinen
04-24-2020, 07:34 AM
Gedmatch could remove those samples if you upload too much artificial genomes. They were already attacked for letting such possibility.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6946563/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6992384/
I could write them to declare the situation. I use only public study data and my samples serve Gedmatch users after I have published them. There is no private issues and only advantages.
the north-euro component in dodecad is based on Lithuanians, so its normal that russians get high.
Dude, I didn't claim otherwise. I know more about my group than anyone else on this forum. Your comment was unnecessary.
Just fucking read this once again
Most of us are 55-60%. Some are 61-62% or so but 66% is definitely super high for an average. I've seen a couple of hundreds of samples, so I can speak confidently about it.
Lemminkäinen
04-24-2020, 07:40 AM
You got me wrong. I just told you there are updated spreadsheets if you want to check them.
Of course it's genetically Russian, one of the oldest Russian cities. A sample size of 2 is way too small to make any conclusions. But I do know that Pskov is heavily Baltic, no news for me.
All right, pardon me. I can't use spread sheets, I need million SNPs or so. Those two samples are however from a published study, obviously released by Russian scientists
All right, pardon me. I can't use spread sheets, I need million SNPs or so. Those two samples are however from a published study, obviously released by Russian scientists
No problem. I'm looking forward to new averages by the way!
Lemminkäinen
04-24-2020, 08:21 AM
I sent an approval request to Gedmatch to make this official. I don't add more samples before their answer and principles.
Hey, sir, why did you delete the kits? I wanted to add a couple of them to Dodecad K12b Vahaduo. Just post Latvian and Estonian if you don't them to be public.
Powered by vBulletin® Version 4.2.3 Copyright © 2025 vBulletin Solutions, Inc. All rights reserved.