# Thread: Ancient K=9 G25 calculator

1. Originally Posted by michal3141
How would you approach trying to find a set of populations (as basic ones as possible: EHG, CHG, WHG, Levant_N, Iran_N, Barcin_N etc.) to minimize the distance?

I am not completely happy with the results below:

[1] "distance%=6.4918"

michal3141

EHG,44.8
Barcin_N,38.2
WHG,7.4
CHG,5.8
Levant_N,3.8

What population/s am I missing?
That method always leaves much to be desired, for me anyway.

I can get my distance quite low using it, but being Eastern European you may inflate your EHG without the use of Northeastern European hunter gatherers. I've got a pretty good spreadsheet I have been using for ancient fits and I always include Narva Lithuania. If the fit is still bad, just add in Yamnaya.

2. ## The Following User Says Thank You to LTG For This Useful Post:

michal3141 (11-26-2018)

3. Originally Posted by michal3141
In your spreadsheet which populations were used for Anatolia_N, Iran_N, East Asian?

I guess SSA = Yoruba, AASI = Onge. Is this correct?

If we use Yamnaya_Samara to replace both EHG and CHG my distance gets a bit lower but not much:

[1] "distance%=6.0342"

michal3141

Yamnaya_Samara,53.6
Barcin_N,24.2
WHG,16.2
Levant_N,6

For some reason introducing Baltic_BA instead of Yamnaya_Samara helps a lot:

[1] "distance%=3.8731"

michal3141

Baltic_BA,74.4
Barcin_N,11.4
Levant_N,7.4
Ganj_Dareh_N,4.8
Han,1
Paniya,1
Yes, I mean I can get it low enough for most people there, but I was trying to use "base" components, which are older than the steppe migrations.
I was also reluctant in adding an Anatolia Neolithic component, but I did it after the release of this paper: http://eurogenes.blogspot.com/2018/0...mers-were.html
Prior to that, I just had considered Levant_N as a "pure" (no WHG) proxy for Anatolia_N.

Regarding the populations, I'm using Barcin Neolithic for Anatolia_N, Ganji Dareh for Iran_N and Han_N China (but I think I replaced it with Mongola for the Turkics, as the distance with Han was ridiculous).
You are right about the others.

4. ## The Following 2 Users Say Thank You to Dorkymon For This Useful Post:

JMcB (11-29-2018),  michal3141 (11-26-2018)

5. "1. CLOSEST SINGLE ITEM DISTANCE%"
Beaker_Iberia Armenia_ChL Baltic_BA Anc_Saami Levant_N
9.449506 11.396843 13.910164 22.222077 23.011851
Iran_N South_Asia East_Asia
30.260101 43.660169 60.647067

"2. FULL TABLE nMONTE"
"penalty= 0"

"distance%=3.1162"

Baltic_BA,37.2
Armenia_ChL,30.8
Beaker_Iberia,21.2
Levant_N,10.2
East_Asia,0.6

"2. FULL TABLE nMONTE"
"penalty= 0.001"

"distance%=3.4068"

Beaker_Iberia,33.2
Baltic_BA,30.2
Armenia_ChL,26.4
Levant_N,8.2
Iran_N,1.6
Anc_Saami,0.2
East_Asia,0.2

6. ## The Following 4 Users Say Thank You to dink For This Useful Post:

Просигој (11-29-2018),  Hurricane (11-29-2018),  JMcB (11-29-2018),  michal3141 (11-29-2018)

7. pen=0.001

distance%=4.8317

Ais_scaled

Beaker_Iberia,55
Baltic_BA,26
Armenia_ChL,12.4
Anc_Saami,5.8
Iran_N,0.8

pen=0

distance%=4.8179

Ais_scaled

Beaker_Iberia,55.6
Baltic_BA,25.2
Armenia_ChL,10.4
Anc_Saami,6.6
Iran_N,2.2

Not much difference with pen=0, except some of my Armenia_ChL goes to Iran_N.

8. ## The Following 4 Users Say Thank You to Ais For This Useful Post:

dink (11-29-2018),  Hurricane (11-29-2018),  JMcB (11-29-2018),  michal3141 (11-29-2018)

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•