PDA

View Full Version : Mapping Your Location (GEDmatch calculators)



DMXX
12-27-2013, 07:58 PM
Some of the Oracle results people get tend to make sense but lack a context to which they can be interpreted. This is especially true when users examine results from several calculators.

I attempted to convene this assortment of information by taking one Oracle result from four calculators (MDLP World-22, Eurogenes K13, Dodecad K12b and HarappaWorld) to see what sort of picture is painted on a map.

This was done by visualising the genetic cline between proposed donor populations A and B and marking the point indicated by the Oracle program, repeating the process with other results, then creating a bounded area indicating likely area of ancestry. I've also charted my physical ancestry and where I should hypothetically land. The main caveat here came with the placement of the donor populations on the map, particularly with the Iranians, given they are mixed from across the country. I decided to place the Tajiks to the west (most likely sampling site) and the Iranians in the middle. The resulting image is shown below. You can also view the raw data (as well as the chosen Oracle results in bold) at the very bottom, as well as the step-by-step guide on how to create something similar.


http://img827.imageshack.us/img827/8338/xecy.png

Summary of the results:


I consistently emerge as either mostly Iranian with minor European admixture or a Near-Eastern and Tajik combination on all tests.
The area bound by the four Oracle results is in the centre-north of the Iranian plateau. This is significantly east of the presumed overall location based on the two points of physical ancestral origin.
The most likely explanation is that my paternal side (Azeri Iranian) has a similar genetic constitution as my maternal (Persian) and most of my paternal ancestors expanded westwards from a common area. So, despite being from West Azerbaijan province and speaking Azeri as a mother-tongue, they will be similar to other north Iranians, which I'm not surprised to see. Incidentally, the same explanation applies to the Kurds, who generally sit in Northwest Iran despite being positioned even further west. The Iranic speakers (plus Azeris) inhabiting that horizontal stretch of land from Anatolia to Mazandaran are therefore likely to be of the same general genetic stock (there will of course be some differences, albeit minor).
Additionally, in support of the above, my maternal side's physical area of ancestry is well within the bounded area formed by the Oracle results.
One important point to raise is that the position of the Iranians would have made the bounded area more specific and probably localised to around the Alborz mountains beneath the Caspian sea.


Reassuringly, this position corresponds perfectly with Humanist's Indo-European Language SPA map (http://www.anthrogenica.com/showthread.php?1677-Indo-European-Language-SPA-Map):


http://i1096.photobucket.com/albums/g326/dok101/Faces/SPA_Indo_European.jpg

Any comments would be welcomed, as would other members exploring their Oracle results in a similar manner here.

Admixture Calculator Oracle results (criteria = all populations present on map, no diaspora populations, lowest genetic distance possible)



MDLP World-22
# Primary Population (source) Secondary Population (source) Distance
--->1 53.6% Druze (derived) + 46.4% Tadjik (derived) @ 3.31<---
2 56.2% Lebanese (derived) + 43.8% Tadjik (derived) @ 3.78
3 55.5% Jew_Georgia (derived) + 44.5% Tadjik (derived) @ 4.28
4 83.6% Turk (derived) + 16.4% Burusho (derived) @ 4.29
5 54.9% Jew_Syria (derived) + 45.1% Pashtun (derived) @ 4.32
6 82.5% Turk (derived) + 17.5% Pathan (derived) @ 4.37
7 63.8% Iranian (derived) + 36.2% Tatar_Crim (derived) @ 4.39
8 89.8% Turk (derived) + 10.2% Hindu (derived) @ 4.51
9 85.5% Turk (derived) + 14.5% Sindhi (derived) @ 4.57
10 82.3% Turk (derived) + 17.7% Balochi (derived) @ 4.58

Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 92.6% Iranian + 7.4% Southeast_English @ 3.04
2 92.7% Iranian + 7.3% Danish @ 3.04
3 92.1% Iranian + 7.9% German @ 3.06
4 91.1% Iranian + 8.9% Austrian @ 3.07
5 92.8% Iranian + 7.2% Orcadian @ 3.08
6 92.9% Iranian + 7.1% North_Swedish @ 3.09
7 92.9% Iranian + 7.1% Norwegian @ 3.09
--->8 91.1% Iranian + 8.9% Hungarian @ 3.1<---
9 92.8% Iranian + 7.2% Swedish @ 3.1
10 92.6% Iranian + 7.4% Dutch @ 3.12

Dodecad K12b
# Primary Population (source) Secondary Population (source) Distance
1 69.7% Uzbekistan_Jews (Behar) + 30.3% Tajiks (Yunusbayev) @ 2.58
2 59.7% Azerbaijan_Jews (Behar) + 40.3% Tajiks (Yunusbayev) @ 2.64
--->3 58.7% Assyrian (Dodecad) + 41.3% Tajiks (Yunusbayev) @ 2.67<---
4 59.6% Georgia_Jews (Behar) + 40.4% Tajiks (Yunusbayev) @ 2.86
5 94.6% Iranian (Dodecad) + 5.4% Ukranians (Yunusbayev) @ 3.11
6 95% Iranian (Dodecad) + 5% Russian_B (Behar) @ 3.13
7 94.9% Iranian (Dodecad) + 5.1% Mixed_Slav (Dodecad) @ 3.14
8 95% Iranian (Dodecad) + 5% Russian (Dodecad) @ 3.15
9 95.1% Iranian (Dodecad) + 4.9% Belorussian (Behar) @ 3.15
10 94.9% Iranian (Dodecad) + 5.1% Polish (Dodecad) @ 3.16

HarappaWorld
# Primary Population (source) Secondary Population (source) Distance
1 92.3% iranian (harappa) + 7.7% ashkenazi (harappa) @ 1.44
2 92% iranian (harappa) + 8% ashkenazy-jew (behar) @ 1.5
3 94% iranian (harappa) + 6% tuscan (1000genomes) @ 1.5
4 95% iranian (harappa) + 5% italian (hgdp) @ 1.5
5 96% iranian (harappa) + 4% spaniard (1000genomes) @ 1.54
6 94.6% iranian (harappa) + 5.4% romanian-a (behar) @ 1.6
--->7 94.5% iranian (harappa) + 5.5% bulgarian (yunusbayev) @ 1.61<---
8 96.1% iranian (harappa) + 3.9% french (hgdp) @ 1.62
9 96.8% iranian (harappa) + 3.2% spain-basc (henn2012) @ 1.69
10 96.9% iranian (harappa) + 3.1% basque (hgdp) @ 1.7





1. Visit GEDmatch (http://gedmatch.com/) (presuming you already have genomic raw data and have an account, otherwise, acquire these)
2. Select Admixture on the main menu.
3. Insert your ID and select any of the open-source projects (MDLP, Eurogenes, Dodecad, HarappaWorld) and then any of the calculators.
4. Once the calculator has finished, click the Oracle button (not Oracle-4) underneath your generated component scores.
5. Scroll down to the bottom and inspect the "Mixed Mode Population Sharing:" results. Pick one, preferably one with the lowest genetic difference (GD) to ensure better accuracy and one which includes non-diaspora/recently non-admixed populations (localising Ashkenazi Jewish or African Americans as a donor population on a map will be difficult due to subjective guidance regarding their placement on a map).
6. Repeat the above with at least two other calculators and keep note of the results. For a minimalist approach, Europeans are better off using Eurogenes, Dodecad and MDLP. South Asians are recommended to have HarappaWorld included. Those from elsewhere in the world are free to use any combination, as none of these are specific for other regions.
7. Download this map (http://commons.wikimedia.org/wiki/File:A_large_blank_world_map_with_oceans_marked_in _blue.PNG) (from Wikipedia) or the map below (for McDonald BGA version) and Paint.Net (http://paint.net/) (open-source image editor). Feel free to use another editing software. I prefer Paint.Net because it indicates the 1/3 increments along any line drawn.
8. Open the map with Paint.Net/another image editor. Pinpoint your McDonald BGA average spot or physical ancestral location if desired.
9. With a colour specific to the open-source calculator you're going to use, pinpoint the location where each donor population for your selected Oracle result comes from. If uncertain, look up roughly where they're from (e.g. Pakistani Pashtuns will be around NW Pakistan close to the Afghan border). If a national average (e.g. German_Dodecad), place in the middle of the country.
10. Draw a line between both donor populations. Estimate where on the line you'll fall. Note the numbers are flipped round in practice; for instance, if the Oracle is 70% German + 30% Ukrainian, the spot will end up around the 30% mark on the German end. Make a spot on the line wherever this may be.
11. Repeat steps 9+10 for all the other Oracle runs, remembering to use different colours for the calculators to keep track.
12. Join these spots together with a different coloured line, forming the "bounded area" where your ancestry can be narrowed down from.
13. Completed. Make all the relevant inferences from the results, compare to the additional data in step 8 if present.


2364

AJL
12-27-2013, 08:31 PM
You can do something similar with Population Finder purely algebraically as it were, essentially by inputting the populations (I use their capital cities) and their respective weights, then getting a GPS coordinate.

For example, my own PF:

Continent (Subcontinent) Population Percentage MOE
Europe Finnish, French, Orcadian, Romanian, Russian 40.62% ±26.27%
Europe (Southern European) Sardinian, Tuscan 59.38% ±26.27%

Then I derive lat and long and paste into Excel:

CITY LAT LONG WEIGHT
Helsinki 60.207 24.917 0.08124
Moscow 55.75 37.6167 0.08124
Paris 48.908 2.329 0.08124
Bucharest 44.418 26.082 0.08124
Lerwick 60.163 -1.1 0.08124
AVERAGE 53.8892 17.96894 0.4062 Karsin, Poland

Cagliari 39.232 9.141 0.2969
Florence 43.866 11.162 0.2969
AVERAGE 41.549 10.1515 0.5938 Tyrhennian Sea

WEIGHTED 46.56158924 13.32694413 1.000 Möderndorf, Austria

This is a reasonable average of my known ethnicities.

Humanist
12-27-2013, 10:27 PM
Reassuringly, this position corresponds perfectly with Humanist's Indo-European Language SPA map (http://www.anthrogenica.com/showthread.php?1677-Indo-European-Language-SPA-Map)

Would like to add that these are the median coordinates for Indo-European speaking populations from a Eurogenes SPA run (excluding Indians and Pakistanis) from 2012.

vettor
12-27-2013, 10:58 PM
You can do something similar with Population Finder purely algebraically as it were, essentially by inputting the populations (I use their capital cities) and their respective weights, then getting a GPS coordinate.

For example, my own PF:

Continent (Subcontinent) Population Percentage MOE
Europe Finnish, French, Orcadian, Romanian, Russian 40.62% ±26.27%
Europe (Southern European) Sardinian, Tuscan 59.38% ±26.27%

Then I derive lat and long and paste into Excel:

CITY LAT LONG WEIGHT
Helsinki 60.207 24.917 0.08124
Moscow 55.75 37.6167 0.08124
Paris 48.908 2.329 0.08124
Bucharest 44.418 26.082 0.08124
Lerwick 60.163 -1.1 0.08124
AVERAGE 53.8892 17.96894 0.4062 Karsin, Poland

Cagliari 39.232 9.141 0.2969
Florence 43.866 11.162 0.2969
AVERAGE 41.549 10.1515 0.5938 Tyrhennian Sea

WEIGHTED 46.56158924 13.32694413 1.000 Möderndorf, Austria

This is a reasonable average of my known ethnicities.

we might be talking about the same thing, but it was explained to me as an example of yours.........from the % of PF, ie, sardinian capital, run 40.62% of the kilometres towards helsinki capital and plot mark, do the same for sardinain to moscow, paris, bucharest and lerwick...................then do the same for flaorence and the helsinki , moscow,paris, bucahest and lerwick .
once all are plot , that area where all your plots are is you starting area.

I did that with mine and it was about 30k difference with what doug plotted me

DMXX
12-27-2013, 11:31 PM
Our Azeri member Imadaddin Nasimi asked me to complete a similar chart based on several of his Oracle scores, and he was gracious enough to let me share it with the community.

The most striking feature of his results here are the maintained ties with Central Asian Turkish speaking ethnic groups (Uyghur, Turkmen, Uzbek). With two of the three Oracle scores, a majority Armenian component was found. I commented that this was how I would've imagined Azeris from the Republic to be; predominantly south Caucasian with significant ancestry from Central Asia, clearly from Oghuz Turkish heritage. In Imadaddin's case, it appears to be at least 20% if the Central Asian Turkish values are to be interpreted literally. The net result is a bounded area over the Caspian Sea. The McDonald BGA average spot is also shown in green.


http://img593.imageshack.us/img593/5827/2ax2.png

Wulf Talented
12-27-2013, 11:37 PM
I've always wanted to try this out, when I first posted my FTDNA family finder results on another site a forumer there said if I ran a line from the Middle Eastern population towards London at 93.73% Km ( something like that ), I'd find my location? Somewhere around Kent/Netherlands?

FTDNA results
Continent (Subcontinent) Population Percentage Margin of Error
Europe (Western European) Orcadian 93.73% ±1.91%
Middle East Palestinian, Adygei, Bedouin, Bedouin South, Druze, Iranian, Jewish, Mozabite 6.27% ±1.91%

Or is this sort of thing exclusively for Gedmatch Admix calculators? Cause I'd love to do that too!

DMXX
12-27-2013, 11:47 PM
Or is this sort of thing exclusively for Gedmatch Admix calculators? Cause I'd love to do that too!

Anyone can calculate this sort of thing; it's literally just establishing both donor populations which constitute "you" genetically and then determining where you fall on that artificial cline based on your values.

I had a go at yours, with "European" set on Orkney and "Middle-Eastern" on central Iraq. You end up in the upper-left quadrant of the North Sea. The only conceivable way you could end up around the Netherlands was if one of the donor sites was set further south (e.g. moving your "European" to the Midlands somewhere).

Wulf Talented
12-27-2013, 11:53 PM
Anyone can calculate this sort of thing; it's literally just establishing both donor populations which constitute "you" genetically and then determining where you fall on that artificial cline based on your values.

I had a go at yours, with "European" set on Orkney and "Middle-Eastern" on central Iraq. You end up in the upper-left quadrant of the North Sea. The only conceivable way you could end up around the Netherlands was if one of the donor sites was set further south (e.g. moving your "European" to the Midlands somewhere).

Upper left quadrant of the North Sea? That's quite interesting, thanks though! If I gave you my oracle results, would you be interested in having fun with those? I'd be happy for you to post the results public too.

Edit, or would there be no difference between Oracle results and the FTDNA population reference?

DMXX
12-28-2013, 12:06 AM
Oracle results would be preferred as they're more specific. I had a hard time deciding where to place the Near Eastern reference site in that above attempt.

You're more than welcome to PM me your Oracle results from several calculators.

Wulf Talented
12-28-2013, 12:09 AM
Oracle results would be preferred as they're more specific. I had a hard time deciding where to place the Near Eastern reference site in that above attempt.

You're more than welcome to PM me your Oracle results from several calculators.

Certainly, They'll be with you shortly!

Anglecynn
12-28-2013, 02:02 AM
This is interesting, i've tried it with my own results but not with any of my other family results, although it was rather crude. I usually end up in a triangle between southeast England, northern France, Belgium and the Netherlands using just random places in countries rather than their genetic position.

I've made some crude ones with the EUtest V2 and the map that Humanist made, they sort of make sense, but i'm not sure how to exactly measure the distances on something like paint.

AJL
12-28-2013, 02:16 AM
we might be talking about the same thing, but it was explained to me as an example of yours.........from the % of PF, ie, sardinian capital, run 40.62% of the kilometres towards helsinki capital and plot mark, do the same for sardinain to moscow, paris, bucharest and lerwick...................then do the same for flaorence and the helsinki , moscow,paris, bucahest and lerwick .
once all are plot , that area where all your plots are is you starting area.

I did that with mine and it was about 30k difference with what doug plotted me

Yes, my way is a quick mathematical way to do the same (for those who are geographically challenged!) and ends up with results close to McDonald, who put me near Torino. The margin of error is so high for me, though, that I could be pretty well anywhere in south-central Europe.

Anglecynn
12-28-2013, 02:24 AM
This is interesting, i've tried it with my own results but not with any of my other family results, although it was rather crude. I usually end up in a triangle between southeast England, northern France, Belgium and the Netherlands using just random places in countries rather than their genetic position.

I've made some crude ones with the EUtest V2 and the map that Humanist made, they sort of make sense, but i'm not sure how to exactly measure the distances on something like paint.

If anyone knows a more accurate way for me to do it, would they be able to help me out?

AJL
12-28-2013, 02:26 AM
Oracle results would be preferred as they're more specific. I had a hard time deciding where to place the Near Eastern reference site in that above attempt.

You're more than welcome to PM me your Oracle results from several calculators.

That's a good point with nonspecific multiple populations close to each other.

My top-10 Dodecad V3 Oracle results are


1 60.1% N._European (Xing) + 39.9% Sephardic_Jews (Behar) @ 2.4
2 52.7% Sephardic_Jews (Behar) + 47.3% Swedish (Dodecad) @ 2.42
3 79% Tuscan (Henn) + 21% FIN (1000Genomes) @ 2.63
4 59.1% German (Dodecad) + 40.9% Sephardic_Jews (Behar) @ 2.79
5 54.4% German (Dodecad) + 45.6% Morocco_Jews (Behar) @ 3.07
6 58.5% Argyll (1000 Genomes) + 41.5% Sephardic_Jews (Behar) @ 3.14
7 81.6% Tuscan (Henn) + 18.4% Finnish (Dodecad) @ 3.2
8 56% CEU (HapMap) + 44% Morocco_Jews (Behar) @ 3.31
9 52.2% Ashkenazy_Jews (Behar) + 47.8% CEU (HapMap) @ 3.45
10 53.6% Sephardic_Jews (Behar) + 46.4% Norwegian (Dodecad) @ 3.47


This still always ends up around the Alps.

AJL
12-28-2013, 02:26 AM
If anyone knows a more accurate way for me to do it, would they be able to help me out?

You could try my technique, if you have specific locales and don't mind looking up latitudes and longitudes and have a spreadsheet program.

Anglecynn
12-28-2013, 02:29 AM
You could try my technique, if you have specific locales and don't mind looking up latitudes and longitudes and have a spreadsheet program.

Ok i'll give it a go tommorow, it did look very accurate.

Anglecynn
12-28-2013, 02:47 AM
I tried today, ended up falling out of the ferry by the looks of it. That's using the Population Finder. If i could find the exact coordinates for the populations on Humanist's map i could try it with a Eurogenes calculator. I think i did it right, as I've tried something similar with the Population Finder data before and it had similar results. I think it's because if Orkney is used, they don't actually cluster on the bit of land that they live on, if that makes sense. Logically speaking, i should be somewhere like Belgium.:

http://i628.photobucket.com/albums/uu7/Brodir93/PopulationFinder.png

AJL
12-28-2013, 02:56 AM
Yes it looks like you descend from a particularly intelligent line of dolphin there! I suspect if you used Crewe or Hull or somewhere sort of mid-northish instead of the actual Orkneys, you would hit land somewhere (though plenty of people still end up swimming in the English Channel).

MitchellSince1893
12-28-2013, 04:49 AM
One might consider using the population centroid for a country. This website contains European ones. Scroll to the bottom for the most recent 2005 data. http://cs.ecs.baylor.edu/~hamerly/software/europe_population_weighted_centers.txt

Remember the one for Ireland is for the Republic of Ireland. For the whole Island of Ireland consider using near Delvin, Ireland

MitchellSince1893
12-28-2013, 08:39 AM
When I use gedmatch, MDLP I end up in SW Germany.

With Dodecad I end up in Western Germany, The Netherlands, Belgium, and in the North Sea.

With Eurogenes I'm in the North Sea, Northern Germany, Denmark, and Norway.

I'm tempted to adopt Heligoland as my homeland :)

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

DMXX
12-28-2013, 01:16 PM
Here's Wulf Talented and Sein's results, with the original data used for each:


http://img46.imageshack.us/img46/1355/v9lr.png

Looks to be mostly British, but something's pulling you more towards continental Europe. This is precisely why I enjoy this method; it gives you a global view of your data. If you only viewed your MDLP result, one would begin searching for a Near-Eastern cause for the shift.



Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 88.9% Irish + 11.1% Serbian @ 1.53
MDLP World-22
8 96% CEU_V (derived) + 4% Druze (derived) @ 0.84
Dodecad K12b
2 61.2% British (Dodecad) + 38.8% Hungarians (Behar) @ 1.37



http://imageshack.us/a/img42/9797/5vpu.png

Generally slightly more western and northern than the presumed zone of origin around the Durand line. Slightly more north than McDonald's average spot.



MDLP World-22
1 91.2% Pathan (derived) + 8.8% Palestinian (derived) @ 3.86
Eurogenes K13
1 88.8% Pathan + 11.2% Abhkasian @ 2.73
Dodecad K12b
1 61.6% Jatt (Dodecad) + 38.4% Turkmens (Yunusbayev) @ 4.72
HarappaWorld
[1,] "59.6% punjabi_harappa_10 + 40.4% turkmen_yunusbayev_11" "4.0437"


If I was blessed with the luxury of time this holiday, there would've been no restriction in the number of maps I can produce. Unfortunately I'm running very low on spare time these days. For those interested in creating similar maps, I've written a step-by-step guide on how to make them below. If a rough geographic spot is all that's desired rather than an overall view of ancestral pulls, AJL's method (http://www.anthrogenica.com/showthread.php?1778-Mapping-Your-Location-(GEDmatch-calculators)&p=24752&viewfull=1#post24752) makes an excellent alternative (probably takes less time as well). I have also entered this information into the original post.



1. Visit GEDmatch (http://gedmatch.com/) (presuming you already have genomic raw data and have an account, otherwise, acquire these)
2. Select Admixture on the main menu.
3. Insert your ID and select any of the open-source projects (MDLP, Eurogenes, Dodecad, HarappaWorld) and then any of the calculators.
4. Once the calculator has finished, click the Oracle button (not Oracle-4) underneath your generated component scores.
5. Scroll down to the bottom and inspect the "Mixed Mode Population Sharing:" results. Pick one, preferably one with the lowest genetic difference (GD) to ensure better accuracy and one which includes non-diaspora/recently non-admixed populations (localising Ashkenazi Jewish or African Americans as a donor population on a map will be difficult due to subjective guidance regarding their placement on a map).
6. Repeat the above with at least two other calculators and keep note of the results. For a minimalist approach, Europeans are better off using Eurogenes, Dodecad and MDLP. South Asians are recommended to have HarappaWorld included. Those from elsewhere in the world are free to use any combination, as none of these are specific for other regions.
7. Download this map (http://commons.wikimedia.org/wiki/File:A_large_blank_world_map_with_oceans_marked_in _blue.PNG) (from Wikipedia) and Paint.Net (http://paint.net/) (open-source image editor). Feel free to use another editing software. I prefer Paint.Net because it indicates the 1/3 increments along any line drawn.
8. Open the map with Paint.Net/another image editor. Pinpoint your McDonald BGA average spot or physical ancestral location if desired.
9. With a colour specific to the open-source calculator you're going to use, pinpoint the location where each donor population for your selected Oracle result comes from. If uncertain, look up roughly where they're from (e.g. Pakistani Pashtuns will be around NW Pakistan close to the Afghan border). If a national average (e.g. German_Dodecad), place in the middle of the country.
10. Draw a line between both donor populations. Estimate where on the line you'll fall. Note the numbers are flipped round in practice; for instance, if the Oracle is 70% German + 30% Ukrainian, the spot will end up around the 30% mark on the German end. Make a spot on the line wherever this may be.
11. Repeat steps 9+10 for all the other Oracle runs, remembering to use different colours for the calculators to keep track.
12. Join these spots together with a different coloured line, forming the "bounded area" where your ancestry can be narrowed down from.
13. Completed. Make all the relevant inferences from the results, compare to the additional data in step 8 if present.

Wulf Talented
12-28-2013, 03:22 PM
Would the pull towards the east be my Polish ancestry? Or is that pull indicating Jewish ancestry?

Thanks very much for doing this, it's really interesting!

DMXX
12-28-2013, 03:23 PM
Would the pull towards the east be my Polish ancestry? Or is that pull indicating Jewish ancestry?

If you have reason to suspect Jewish ancestry it could well be that, although I'm not familiar with your background to comment any more specifically. Polish ancestry would quite logically pull you toward continental Europe.

Wulf Talented
12-28-2013, 03:26 PM
If you have reason to suspect Jewish ancestry it could well be that, although I'm not familiar with your background to comment any more specifically. Polish ancestry would quite logically pull you toward continental Europe.

Well I never received any Ashkenazi ancestry on my 23andMe test, and I only have one Ashkenazi match out of a dozen or so Polish matches, I suspect I match her due to her Polish ancestry and not the Ashkenazi side. My Polish ancestor was a Catholic too, but I wasn't ruling out the possibility considering he came from Warsaw. The Jtest only gives me some 3.08% Ashkenazi result, which according to Davidski is likely to be minor distant or none at all.

So it's likely to be just the Polish ancestry, I suspect. I only asked if it might be Jewish because one of the populations mentions Druze at 4%.

Dr_McNinja
12-28-2013, 04:34 PM
You can do something similar with Population Finder purely algebraically as it were, essentially by inputting the populations (I use their capital cities) and their respective weights, then getting a GPS coordinate.

For example, my own PF:

Continent (Subcontinent) Population Percentage MOE
Europe Finnish, French, Orcadian, Romanian, Russian 40.62% ±26.27%
Europe (Southern European) Sardinian, Tuscan 59.38% ±26.27%

Then I derive lat and long and paste into Excel:

CITY LAT LONG WEIGHT
Helsinki 60.207 24.917 0.08124
Moscow 55.75 37.6167 0.08124
Paris 48.908 2.329 0.08124
Bucharest 44.418 26.082 0.08124
Lerwick 60.163 -1.1 0.08124
AVERAGE 53.8892 17.96894 0.4062 Karsin, Poland

Cagliari 39.232 9.141 0.2969
Florence 43.866 11.162 0.2969
AVERAGE 41.549 10.1515 0.5938 Tyrhennian Sea

WEIGHTED 46.56158924 13.32694413 1.000 Möderndorf, Austria

This is a reasonable average of my known ethnicities.I tried this method and got this area:

http://i.imgur.com/CCCoYwk.png

My PF results:


North India - 82.36%
Palestinian, Adygei, Druze, Iranian, Jewish - 8.51%
Finnish, Russian, Orcadian - 9.13%

Táltos
12-28-2013, 04:53 PM
Thanks DMXX for posting how to do this, but I'm kind of confused how to still make it work. :embarrassed: I never was able to figure out how to make SPA work. I'm okay with downloading the raw data for this stuff, and upload to Gedmatch and to send Dr. McDonald the data. Just when it comes to other cool tools like this I go Doh! I know you said you are busy and wouldn't be able to do anymore maps. If anyone else out there is proficient in this, and if you have time could someone please make a map for me? Thanks. I decided to use the all the World calculators, and for Eurogenes I just picked the one with the lowest distance. Eurogenes always gives me higher distances than the other calculators. I also added my second lowest distance on MDLP World22. I wasn't sure that my first one would be valid enough with the North European Mesolithic. I guess you could place that almost anywhere in Northern Europe?

MDLP World-22
1 98.3% Bosnian (derived)+ 1.7% North-European-Mesolithic (ancestral)@ 0.99
2 97.7% Bosnian (derived) +2.3% Saami (derived)@1.01

MDLP World
1 52.5% Gagauz + 47.5% Polish_V @ 0.89

Dodecad World9
1 85.6% German (Dodecad) + 14.4 Georgia_Jews (Behar)@0.56

Eurogenes K13
1 51.5% Polish + 48.5% Romanian @1.54

Dr_McNinja
12-28-2013, 05:01 PM
I tried this method and got this area:

http://i.imgur.com/CCCoYwk.png

My PF results:
North India - 82.36%
Palestinian, Adygei, Druze, Iranian, Jewish - 8.51%
Finnish, Russian, Orcadian - 9.13%I used New Delhi for North India there. If I use Lucknow instead (right in the middle of North India/Uttar Pradesh), it comes out VERY close to where my family's village is:

http://i.imgur.com/DZnqZL7.png

That star is where my family's village is (Toba Tek Singh). Most of my family hails from over the border in India so it's remarkable that it winds up being this close.

EDIT: These are the locations I used:


Lucknow - 26.847 N 80.947 E - 0.8236

Jerusalem - 31.7833 N 35.2167 E - .40
Maykop (Aydgei) - 44.6 N 40.0833 E - .20
Damascus - 33.513 N 36.292 E - .20
Tehran - 35.6961 N 51.4231 E - .20
(Total - 0.0851)

Helsinki - 60.1708 N 24.9375 E - .33
Moscow - 55.75 N 37.6167 E - .33
Orkney Islands - 59 N, 3 W - .33
(Total - 0.0913)Still amazed how well FTDNA's Population Finder worked for me. As good as McDonald's analysis. Or at least it seems to function on a similar methodology? I know I should have more Western European than Eastern European admixture and more Caucasus (Adygei) than Mideastern admixture, but they use the right amounts of all of them to arrive at a similar place. And I had to guess a location for North India, which is a huge area on their map, even though Lucknow seems like a safe bet. Another liberty I took was using Jerusalem for both Palestinian and Jewish.

vettor
12-28-2013, 05:57 PM
Here's Wulf Talented and Sein's results, with the original data used for each:


http://img46.imageshack.us/img46/1355/v9lr.png

Looks to be mostly British, but something's pulling you more towards continental Europe. This is precisely why I enjoy this method; it gives you a global view of your data. If you only viewed your MDLP result, one would begin searching for a Near-Eastern cause for the shift.



Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 88.9% Irish + 11.1% Serbian @ 1.53
MDLP World-22
8 96% CEU_V (derived) + 4% Druze (derived) @ 0.84
Dodecad K12b
2 61.2% British (Dodecad) + 38.8% Hungarians (Behar) @ 1.37



http://imageshack.us/a/img42/9797/5vpu.png

Generally slightly more western and northern than the presumed zone of origin around the Durand line. Slightly more north than McDonald's average spot.



MDLP World-22
1 91.2% Pathan (derived) + 8.8% Palestinian (derived) @ 3.86
Eurogenes K13
1 88.8% Pathan + 11.2% Abhkasian @ 2.73
Dodecad K12b
1 61.6% Jatt (Dodecad) + 38.4% Turkmens (Yunusbayev) @ 4.72
HarappaWorld
[1,] "59.6% punjabi_harappa_10 + 40.4% turkmen_yunusbayev_11" "4.0437"


If I was blessed with the luxury of time this holiday, there would've been no restriction in the number of maps I can produce. Unfortunately I'm running very low on spare time these days. For those interested in creating similar maps, I've written a step-by-step guide on how to make them below. If a rough geographic spot is all that's desired rather than an overall view of ancestral pulls, AJL's method (http://www.anthrogenica.com/showthread.php?1778-Mapping-Your-Location-(GEDmatch-calculators)&p=24752&viewfull=1#post24752) makes an excellent alternative (probably takes less time as well). I have also entered this information into the original post.

i butchered mine , then again I am not computer savvy, see below results

I have a question, what are you triangulating the 3 results for? and once triangulated it would mean anywhere inside this triangle?

these are my results
1 75.8% NITAL (North-Italian) + 24.2% RUS (Russian) @ 4.73 (MDLF in RED )

1 90.5% North_Italian + 9.5% South_Finnish @ 2.3 ( Eurogenes in GREEN )

1 61% C_Italian (Dodecad) + 39% Dutch (Dodecad) @ 1.03 ( Dodecad in GREEN )

IN BEIGE colour.... ..DOUG's spot for me

http://img34.imageshack.us/img34/3302/v7z.PNG (http://imageshack.us/photo/my-images/34/v7z.PNG/)

Uploaded with ImageShack.us (http://imageshack.us)

AJL
12-28-2013, 05:58 PM
Still amazed how well FTDNA's Population Finder worked for me. As good as McDonald's analysis. Or at least it seems to function on a similar methodology?

Yes, McDonald designed the basic workings of Population Finder and it was then further tweaked by FTDNA's team. I've heard nmore complaints about it than praises, but if you view it as a "spot on the map" rather than a definitive ethnicity mix, it's pretty good. For ethnicity mixes, the various Oracles seem better, as does 23andme.

AJL
12-28-2013, 06:00 PM
i butchered mine , then again I am not computer savvy, see below results

I have a question, what are you triangulating the 3 results for? and once triangulated it would mean anywhere inside this triangle?

these are my results
1 75.8% NITAL (North-Italian) + 24.2% RUS (Russian) @ 4.73 (MDLF in RED )

1 90.5% North_Italian + 9.5% South_Finnish @ 2.3 ( Eurogenes in GREEN )

1 61% C_Italian (Dodecad) + 39% Dutch (Dodecad) @ 1.03 ( Dodecad in GREEN )

IN BEIGE colour.... ..DOUG's spot for me

http://img34.imageshack.us/img34/3302/v7z.PNG (http://imageshack.us/photo/my-images/34/v7z.PNG/)

Uploaded with ImageShack.us (http://imageshack.us)

Those are still relatively consistent. The north-south placement is always much more certain than the east-west.

DMXX
12-28-2013, 06:11 PM
I have a question, what are you triangulating the 3 results for? and once triangulated it would mean anywhere inside this triangle?


That triangulated region (or quadrilateral if four Oracle results are used) are intended to give you an idea of the territory that's mutually agreed by all the Oracles. It's a best-of-3/4-worlds affair.

The region formed by 3-4 Oracle results is the general genetic area your ancestry supposedly comes from, taking into consideration your majority position and any minority pulls in any given direction. In my case and Wulf's, it's slightly eastwards. In Sein's it's slightly northwest etc.

Wulf Talented
12-28-2013, 06:31 PM
That triangulated region (or quadrilateral if four Oracle results are used) are intended to give you an idea of the territory that's mutually agreed by all the Oracles. It's a best-of-3/4-worlds affair.

The region formed by 3-4 Oracle results is the general genetic area your ancestry supposedly comes from, taking into consideration your majority position and any minority pulls in any given direction. In my case and Wulf's, it's slightly eastwards. In Sein's it's slightly northwest etc.


Does that mean my ancestry would be Kent ( majority ) but due to the Polish pull, it labels me over Belgium/Netherlands? Because the bottom line of that triangle is on Kent, and there's where I was born and that's my Father's lineage.

Edit, what if we were to throw in the EUTest and Jtest?

Tolan
12-28-2013, 06:31 PM
This is my map!:\
1116

Compared to my SPA map:
http://genetics.cs.ucla.edu/spa/download.html
1118

Táltos
12-28-2013, 07:10 PM
Ok guys here is my crude attempt at this, please don't laugh. And I'm sure I don't have anything figured in with the percents. Again here would be the oracles I used.
MDLP World-22
1 98.3% Bosnian (derived)+ 1.7% North-European-Mesolithic (ancestral)@ 0.99
2 97.7% Bosnian (derived) +2.3% Saami (derived)@1.01

MDLP World
1 52.5% Gagauz + 47.5% Polish_V @ 0.89

Dodecad World9
1 85.6% German (Dodecad) + 14.4 Georgia_Jews (Behar)@0.56

Eurogenes K13
1 51.5% Polish + 48.5% Romanian @1.54

Here is my attempt.
1119
What do you guys think?

Sein
12-29-2013, 12:21 AM
Thank you very much DMXX!

So I seem to cluster in Central Afghanistan, on the far western edge of the Pashtun region, from Samangan to Ghazni. This is interesting, since I originate from the eastern edge of the Pashtun region. I think I can explain this. My father's tribe are migrants from Kandahar, and arrived at their present location in fairly recent times. After protracted conflict with closely related tribes, they sought pasture in the fertile plains close to the traditional border of Greater India, taking away land from locals, and distributing it among their various septs. About 25% of my mother's ancestry is derived from a nomadic tribe that used to divide their time between Ghazni and Bannu, with origins in the Loya Paktia region of Afghanistan. The rest of her ancestry is derived from eastern Pashtun tribes, ones completely restricted to contemporary Pakistan. Nevertheless, Pashtuns west and east of the Durand Line shouldn't be too different.

Joe B
12-29-2013, 01:04 AM
This does pull me east. Most likely because of the unknown admixture.
http://farm4.staticflickr.com/3824/11614196666_77e216d712_o_d.jpg
MDLP World-22 Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 95.9% Serbian (derived) + 4.1% North-European-Mesolithic (ancestral) @ 1.1
2* 87.8% Montenegrin (derived) + 12.2% Finnish-North (derived) @ 1.27
3 94.8% Serbian (derived) + 5.2% Saami (derived) @ 1.27
4 83.3% CEU_V (derived) + 16.7% Jew-Iraqi (derived) @ 1.28
5 83.6% CEU_V (derived) + 16.4% Jew-Iran (derived) @ 1.29
6 95.3% Serbian (derived) + 4.7% Bra2 (derived) @ 1.29
7 76.3% German-North (derived) + 23.7% Cypriot (derived) @ 1.33
8 76.2% German-North (derived) + 23.8% Jew_Syria (derived) @ 1.37
9 63.3% Gagauz (derived) + 36.7% Swedish_V (derived) @ 1.38
10 82.7% CEU_V (derived) + 17.3% Jew_Georgia (derived) @ 1.45

Eurogenes K13 Oracle - Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 85.6% German + 14.4% Armenian @ 1.98
2 85.7% German + 14.3% Georgian_Jewish @ 2.06
3 83.1% German + 16.9% Turkish @ 2.12
4 85.3% German + 14.7% Kurdish @ 2.3
5 85.9% German + 14.1% Assyrian @ 2.37
6 85.5% German + 14.5% Azeri_Jewish @ 2.37
7 86.7% German + 13.3% Abhkasian @ 2.52
8 86.1% German + 13.9% Georgian @ 2.57
9 83.2% German + 16.8% Kumyk @ 2.62
10 85.6% German + 14.4% Iranian @ 2.71

Dodecad K12b Oracle – Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 69.3% Norwegian (Dodecad) + 30.7% Cypriots (Behar) @ 1.56
2 53.5% Bulgarian (Dodecad) + 46.5% Argyll (1000Genomes) @ 1.64
3 79.3% German (Dodecad) + 20.7% Sephardic_Jews (Behar) @ 1.77
4 60.9% Swedish (Dodecad) + 39.1% S_Italian_Sicilian (Dodecad) @ 2.09
5 54.6% Bulgarian (Dodecad) + 45.4% Orcadian (HGDP) @ 2.09
6 76.3% German (Dodecad) + 23.7% S_Italian_Sicilian (Dodecad) @ 2.09
7 54.5% Bulgarian (Dodecad) + 45.5% Orkney (1000Genomes) @ 2.1
8 56.4% Romanians (Behar) + 43.6% Argyll (1000Genomes) @ 2.14
9 82.3% German (Dodecad) + 17.7% Cypriots (Behar) @ 2.18
10 54.5% Bulgarian (Dodecad) + 45.5% Irish (Dodecad) @ 2.19

HarappaWorld Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 82.7% utahn-white (1000genomes) + 17.3% georgia-jew (behar) @ 1.32
2 82.9% utahn-white (1000genomes) + 17.1% azerbaijan-jew (behar) @ 1.34
3 82.6% utahn-white (1000genomes) + 17.4% armenian (behar) @ 1.44
4 80% utahn-white (1000genomes) + 20% turk (behar) @ 1.55
5 79.7% utahn-white (1000genomes) + 20.3% turk-kayseri (hodoglugil) @ 1.59
6 81.6% utahn-white (hapmap) + 18.4% lebanese-druze (haber) @ 1.6
7* 79.2% orcadian (hgdp) + 20.8% druze (hgdp) @ 1.64
8 79.1% utahn-white (1000genomes) + 20.9% turk-istanbul (hodoglugil) @ 1.64
9 79.6% utahn-white (hapmap) + 20.4% cypriot (behar) @ 1.66
10 82.8% utahn-white (1000genomes) + 17.2% assyrian (harappa) @ 1.67
Thanks DMXX for this useful exercise. Now I get it!
"Free time" may be rare once you enter a residency program. So double thanks!

vettor
12-29-2013, 01:33 AM
This does pull me east. Most likely because of the unknown admixture.
http://farm4.staticflickr.com/3824/11614196666_6c6d4e19d2_o.jpg
MDLP World-22 Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 95.9% Serbian (derived) + 4.1% North-European-Mesolithic (ancestral) @ 1.1
2* 87.8% Montenegrin (derived) + 12.2% Finnish-North (derived) @ 1.27
3 94.8% Serbian (derived) + 5.2% Saami (derived) @ 1.27
4 83.3% CEU_V (derived) + 16.7% Jew-Iraqi (derived) @ 1.28
5 83.6% CEU_V (derived) + 16.4% Jew-Iran (derived) @ 1.29
6 95.3% Serbian (derived) + 4.7% Bra2 (derived) @ 1.29
7 76.3% German-North (derived) + 23.7% Cypriot (derived) @ 1.33
8 76.2% German-North (derived) + 23.8% Jew_Syria (derived) @ 1.37
9 63.3% Gagauz (derived) + 36.7% Swedish_V (derived) @ 1.38
10 82.7% CEU_V (derived) + 17.3% Jew_Georgia (derived) @ 1.45

Eurogenes K13 Oracle - Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 85.6% German + 14.4% Armenian @ 1.98
2 85.7% German + 14.3% Georgian_Jewish @ 2.06
3 83.1% German + 16.9% Turkish @ 2.12
4 85.3% German + 14.7% Kurdish @ 2.3
5 85.9% German + 14.1% Assyrian @ 2.37
6 85.5% German + 14.5% Azeri_Jewish @ 2.37
7 86.7% German + 13.3% Abhkasian @ 2.52
8 86.1% German + 13.9% Georgian @ 2.57
9 83.2% German + 16.8% Kumyk @ 2.62
10 85.6% German + 14.4% Iranian @ 2.71

Dodecad K12b Oracle – Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 69.3% Norwegian (Dodecad) + 30.7% Cypriots (Behar) @ 1.56
2 53.5% Bulgarian (Dodecad) + 46.5% Argyll (1000Genomes) @ 1.64
3 79.3% German (Dodecad) + 20.7% Sephardic_Jews (Behar) @ 1.77
4 60.9% Swedish (Dodecad) + 39.1% S_Italian_Sicilian (Dodecad) @ 2.09
5 54.6% Bulgarian (Dodecad) + 45.4% Orcadian (HGDP) @ 2.09
6 76.3% German (Dodecad) + 23.7% S_Italian_Sicilian (Dodecad) @ 2.09
7 54.5% Bulgarian (Dodecad) + 45.5% Orkney (1000Genomes) @ 2.1
8 56.4% Romanians (Behar) + 43.6% Argyll (1000Genomes) @ 2.14
9 82.3% German (Dodecad) + 17.7% Cypriots (Behar) @ 2.18
10 54.5% Bulgarian (Dodecad) + 45.5% Irish (Dodecad) @ 2.19

HarappaWorld Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 82.7% utahn-white (1000genomes) + 17.3% georgia-jew (behar) @ 1.32
2 82.9% utahn-white (1000genomes) + 17.1% azerbaijan-jew (behar) @ 1.34
3 82.6% utahn-white (1000genomes) + 17.4% armenian (behar) @ 1.44
4 80% utahn-white (1000genomes) + 20% turk (behar) @ 1.55
5 79.7% utahn-white (1000genomes) + 20.3% turk-kayseri (hodoglugil) @ 1.59
6 81.6% utahn-white (hapmap) + 18.4% lebanese-druze (haber) @ 1.6
7* 79.2% orcadian (hgdp) + 20.8% druze (hgdp) @ 1.64
8 79.1% utahn-white (1000genomes) + 20.9% turk-istanbul (hodoglugil) @ 1.64
9 79.6% utahn-white (hapmap) + 20.4% cypriot (behar) @ 1.66
10 82.8% utahn-white (1000genomes) + 17.2% assyrian (harappa) @ 1.67
Thanks DMXX for this useful exercise. Now I get it!
"Free time" may be rare once you enter a residency program. So double thanks!

I was always advised that orkney "capital" was kent for Audna things.
if you changed it , it would sit in your known ancestry area

Joe B
12-29-2013, 02:03 AM
#14 of the HarappaWorld oracle mixed mode would include more of my known ancestry.
14 81.7% british (1000genomes) + 18.3% druze (hgdp) @ 1.94
Thought about plotting Salt Lake City for the utahn-white in HarappaWorld. That would get more of my ancestral areas. Except for my Irish!
:beerchug:
Using Orkney must have seemed like a good idea at one time, not so much now.

Sein
12-29-2013, 04:17 AM
I tried my hand at this, using DMXX's instructions. His is obviously better, but I wanted to see what would happen if I changed a few things. I put the HGDP Pashtuns at Karbogha Sharif, near Kurram Agency. Some coordinates put them at Parachinar, which is basically in Afghanistan. Nobody seems to know the actual location, but it's somewhere around the Kurram-Bangash area, from Pewar in the west, to Hangu in the east. Also, it's possible that the Yunusbayev Turkmen are from Iran, so I tried that. And I put the Punjabi at the Pakistan-India border. I just wanted to see if anything differs significantly. The triangle is what matters.

My attempt:
1121

Very similar, if not identical, to what DMXX provided, Central Afghanistan. Also, his is just significantly cooler, plain and simple. So I'm more happy with his version :).

Also, I tried AJL's instructions, with some results I had for a friend's Supervised Admixture run. Here it is:
1122

Eastern Iran, close to the border with Afghanistan. I would assume that this is a stretch, but based on the Bandari Iranian individual at HAP, at least some Eastern-South Eastern Iranians are genetically very similar to myself. I like AJL's method, it's very quick.

MitchellSince1893
12-29-2013, 06:56 AM
I was always advised that orkney "capital" was kent for Audna things.


Thanks. That little bit of info took a lot of my results out of the North Sea and put them in Benelux area.

MitchellSince1893
12-29-2013, 06:58 AM
Does anyone know a good lat/long for "Mixed Germanic"?

My best guess is somewhere in the North Sea near the English coast.

vettor
12-29-2013, 08:55 AM
Does anyone know a good lat/long for "Mixed Germanic"?

My best guess is somewhere in the North Sea near the English coast.

Mixed Germanic refers to southern Germany.....bavaria, freiburg, old swabian and alemanni lands. Mixed refers to, the old existing various alpine peoples and germans arriving from the north because that the last place the germans arrived in Germany. IIRC , they have been there only, for about 1500 to 1600 years

Palisto
12-29-2013, 09:41 AM
...
5. Scroll down to the bottom and inspect the "Mixed Mode Population Sharing:" results. Pick one, preferably one with the lowest genetic difference (GD) to ensure better accuracy and one which includes non-diaspora/recently non-admixed populations (localising Ashkenazi Jewish or African Americans as a donor population on a map will be difficult due to subjective guidance regarding their placement on a map)
....

In order to predict the biogeographical ancestry I used exactly your approach, however, I used more than 2000 of the best Oracle results and weighted them based on their GD.
Here is it for Dodecad Globe13
http://www.anthrogenica.com/showthread.php?1772-Iranian-Jews-Reconciling-Autosomal-DNA-and-Physical-Appearance&p=24949&viewfull=1#post24949
http://kurdishdna.blogspot.com/2013/12/globe13-calculator-for-genetic-distance.html

Sein
12-29-2013, 11:08 AM
Thank you Palisto!

Using your approach, my spot on the map is 31.34° N, 64.93° E, which is in between Kandahar Shahr and Lashkar Gah. For one of my top Oracle results, I'm at 33.90 N, 67.73°E, which is northwestern Uruzgan.

Yggdrasil
12-29-2013, 11:25 AM
Here´s my attempt. It´s a bit difficult to place the different Russian spots on the blank map I find. The yellow circle is actual ancestry location. Even if I pick the population sets I like the triangle will not touch the home area.

http://www.anthrogenica.com/attachment.php?attachmentid=1128&d=1388313737

The data:


Eurogenes K13
1 74.3% West_Scottish + 25.7% Kargopol_Russian @ 1.09
2 78.5% Orcadian + 21.5% Kargopol_Russian @ 1.29
3 77% West_Scottish + 23% Erzya @ 1.33
4 76.2% Irish + 23.8% Kargopol_Russian @ 1.42
5 80.9% Orcadian + 19.1% Erzya @ 1.43
6 75.3% Irish + 24.7% East_Finnish @ 1.63
7 76.4% Orcadian + 23.6% Estonian @ 1.71
8 81.8% Norwegian + 18.2% Hungarian @ 1.71
9 78.9% Irish + 21.1% Erzya @ 1.71
10 77.8% Orcadian + 22.2% East_Finnish @ 1.73
11 79.3% Danish + 20.7% Kargopol_Russian @ 1.78
12 75.1% Orcadian + 24.9% Finnish @ 1.81
13 68.4% Orcadian + 31.6% Southwest_Finnish @ 1.88
14 65.4% Irish + 34.6% Southwest_Finnish @ 1.88
15 81.7% Norwegian + 18.3% Austrian @ 1.91
16 72.5% Irish + 27.5% Finnish @ 1.91
17 79.6% Orcadian + 20.4% Lithuanian @ 1.95
18 81.7% Danish + 18.3% Erzya @ 1.96
19 84.2% Dutch + 15.8% East_Finnish @ 2.01
20 73.6% West_Scottish + 26.4% East_Finnish @ 2.01

MDLP World22
1 87.9% CEU_V (derived) + 12.1% Udmurd (derived) @ 2.34
2 55.7% French (derived) + 44.3% Russian_North (derived) @ 2.35
3 87.8% CEU_V (derived) + 12.2% Mari (derived) @ 2.37
4 82.5% CEU_V (derived) + 17.5% Komi (derived) @ 2.39
5 63.7% German (derived) + 36.3% Norwegian_V (derived) @ 2.44
6 83.9% CEU_V (derived) + 16.1% Chuvash (derived) @ 2.54
7 75.5% Welsh (derived) + 24.5% Komi (derived) @ 2.63
8 73.6% CEU (derived) + 26.4% Komi (derived) @ 2.73
9 71.1% CEU (derived) + 28.9% Mordovian (derived) @ 2.77
10 91% Swedish (derived) + 9% Turkmen (derived) @ 2.8
11 94.9% German (derived) + 5.1% Gok4 (derived) @ 2.81
12 87.9% CEU_V (derived) + 12.1% Mansi (derived) @ 2.83
13 90.8% Swedish (derived) + 9.2% Tadjik (derived) @ 2.84
14 70.6% British (derived) + 29.4% Komi (derived) @ 2.85
15 70.7% German (derived) + 29.3% CEU (derived) @ 2.88
16 82.2% CEU_V (derived) + 17.8% Tatar (derived) @ 2.94
17 72.4% CEU (derived) + 27.6% Russian_North (derived) @ 2.95
18 64.8% German-South (derived) + 35.2% Russian_North (derived) @ 2.95
19 89.5% Swedish (derived) + 10.5% Nogai (derived) @ 2.96
20 75.4% German (derived) + 24.6% Orcadian (derived) @ 2.99

Dodecad K12b
1 72% Kent (1000Genomes) + 28% Russian (HGDP) @ 1.9
2 72% Kent (1000Genomes) + 28% Mordovians (Yunusbayev) @ 2.03
3 78.8% Dutch (Dodecad) + 21.2% Russian (HGDP) @ 2.08
4 67.5% Cornwall (1000Genomes) + 32.5% Mordovians (Yunusbayev) @ 2.27
5 73.3% CEU30 (1000Genomes) + 26.7% Russian (HGDP) @ 2.29
6 77.9% Mixed_Germanic (Dodecad) + 22.1% Russian (HGDP) @ 2.3
7 72.8% English (Dodecad) + 27.2% Mordovians (Yunusbayev) @ 2.32
8 73.3% CEU30 (1000Genomes) + 26.7% Mordovians (Yunusbayev) @ 2.32
9 72.9% English (Dodecad) + 27.1% Russian (HGDP) @ 2.34
10 78.9% Dutch (Dodecad) + 21.1% Mordovians (Yunusbayev) @ 2.34
11 67.6% Cornwall (1000Genomes) + 32.4% Russian (HGDP) @ 2.42
12 60.8% Mordovians (Yunusbayev) + 39.2% Pais_Vasco (1000Genomes) @ 2.45
13 80.5% Mixed_Germanic (Dodecad) + 19.5% FIN30 (1000Genomes) @ 2.51
14 87.5% Norwegian (Dodecad) + 12.5% Turkish (Dodecad) @ 2.56
15 68.2% British (Dodecad) + 31.8% Mordovians (Yunusbayev) @ 2.62
16 72.3% British_Isles (Dodecad) + 27.7% Mordovians (Yunusbayev) @ 2.62
17 82.2% Dutch (Dodecad) + 17.8% Chuvashs (Behar) @ 2.63
18 81.5% Mixed_Germanic (Dodecad) + 18.5% Finnish (Dodecad) @ 2.65
19 87.8% Norwegian (Dodecad) + 12.2% Turks (Behar) @ 2.72
20 65.9% French (HGDP) + 34.1% FIN30 (1000Genomes) @ 2.73

Harappa:
1 65.6% utahn-white (1000genomes) + 34.4% russian (behar) @ 1.84
2 62.1% british (1000genomes) + 37.9% russian (behar) @ 2.02
3 76% n-european (xing) + 24% russian (behar) @ 2.07
4 75.1% n-european (xing) + 24.9% mordovian (yunusbayev) @ 2.13
5 73.6% n-european (xing) + 26.4% belorussian (behar) @ 2.53
6 61.9% utahn-white (hapmap) + 38.1% ukranian (yunusbayev) @ 2.54
7 51.2% british (1000genomes) + 48.8% ukranian (yunusbayev) @ 2.55
8 70.8% utahn-white (hapmap) + 29.2% mordovian (yunusbayev) @ 2.56
9 62% utahn-white (1000genomes) + 38% belorussian (behar) @ 2.62
10 67.9% n-european (xing) + 32.1% ukranian (yunusbayev) @ 2.62
11 64.7% utahn-white (1000genomes) + 35.3% mordovian (yunusbayev) @ 2.62
12 54.1% french (hgdp) + 45.9% lithuanian (behar) @ 2.7
13 55.2% utahn-white (1000genomes) + 44.8% ukranian (yunusbayev) @ 2.74
14 73.3% belorussian (behar) + 26.7% spain-basc (henn2012) @ 2.92
15 50.6% orcadian (hgdp) + 49.4% ukranian (yunusbayev) @ 2.95
16 87.8% n-european (xing) + 12.2% finnish (1000genomes) @ 2.97
17 69.8% russian (behar) + 30.2% spain-basc (henn2012) @ 3
18 74.2% belorussian (behar) + 25.8% basque (hgdp) @ 3.03
19 70.7% russian (behar) + 29.3% basque (hgdp) @ 3.04
20 61.3% british (1000genomes) + 38.7% mordovian (yunusbayev) @ 3.08

Palisto
12-29-2013, 11:55 AM
Thank you Palisto!

Using your approach, my spot on the map is 31.34° N, 64.93° E, which is in between Kandahar Shahr and Lashkar Gah. For one of my top Oracle results, I'm at 33.90 N, 67.73°E, which is northwestern Uruzgan.

So, is 31.34° N, 64.93° E a good proxy for your ancestry; is 33.90 N, 67.73°E better?

Dr_McNinja
12-29-2013, 04:17 PM
I used New Delhi for North India there. If I use Lucknow instead (right in the middle of North India/Uttar Pradesh), it comes out VERY close to where my family's village is:

http://i.imgur.com/DZnqZL7.png

That star is where my family's village is (Toba Tek Singh). Most of my family hails from over the border in India so it's remarkable that it winds up being this close.

EDIT: These are the locations I used:

Still amazed how well FTDNA's Population Finder worked for me. As good as McDonald's analysis. Or at least it seems to function on a similar methodology? I know I should have more Western European than Eastern European admixture and more Caucasus (Adygei) than Mideastern admixture, but they use the right amounts of all of them to arrive at a similar place. And I had to guess a location for North India, which is a huge area on their map, even though Lucknow seems like a safe bet. Another liberty I took was using Jerusalem for both Palestinian and Jewish.Forgot to subtract 3 W of Orkney Islands because it's West, not East, but it doesn't change the result by much, it's still between Multan and Faisalabad, just a little further West.

Solothurn
12-29-2013, 05:56 PM
I nearly always come in as Dutch or SE England on the calculators.

I have 96% Orc and 4% ME on PF and so does my paternal cousin. I wonder if this 'pulls' me farther SE of my actual birth place. BTW most of my paternal ancestry is NE England UK.

AJL
12-29-2013, 06:42 PM
I nearly always come in as Dutch or SE England on the calculators.

I have 96% Orc and 4% ME on PF and so does my paternal cousin. I wonder if this 'pulls' me farther SE of my actual birth place. BTW most of my paternal ancestry is NE England UK.

I have some folk from there too on my mother's side -- mainly Teesdale area as well as a few in County Durham. I do remember reading that parts of Yorkshire had fairly heavy French migration, both Huguenot and in earlier times, and then there were the Flemish weavers. I am not sure if we should look to that to account for a fairly small southeastwards pull in your case but I guess it's one possible take on things.

(Farther south, I also have ancestors from the Isle of Axholme area west of Scunthorpe, which was drained by a Dutch engineer in the early 17th century then settled by Flemish/Dutch/Huguenot immigrants.)

Solothurn
12-29-2013, 07:07 PM
Mainly Durham and Northumberland for me! Although the NE had one of the largest Irish migration outside London, so who knows?

"North east England was the fourth largest centre of Irish migration to England in the nineteenth-century"

On Geno I come in at 19% on Southwest Asian, whereas I think 17% is the British 'norm'!

I am sure I looked on one of the calcs yesterday and saw Arabic as one of our mixes. I cannot find Arabic now so maybe I saw it in my sleep lol :P

My Gran was K1a mtDNA and hadn't thought about it until now, but if 'bottlenecked' in villages many of my lines could also have been K1a!




I have some folk from there too on my mother's side -- mainly Teesdale area as well as a few in County Durham. I do remember reading that parts of Yorkshire had fairly heavy French migration, both Huguenot and in earlier times, and then there were the Flemish weavers. I am not sure if we should look to that to account for a fairly small southeastwards pull in your case but I guess it's one possible take on things.

(Farther south, I also have ancestors from the Isle of Axholme area west of Scunthorpe, which was drained by a Dutch engineer in the early 17th century then settled by Flemish/Dutch/Huguenot immigrants.)

AJL
12-29-2013, 07:41 PM
Mainly Durham and Northumberland for me! Although the NE had one of the largest Irish migration outside London, so who knows?

"North east England was the fourth largest centre of Irish migration to England in the nineteenth-century"

On Geno I come in at 19% on Southwest Asian, whereas I think 17% is the British 'norm'!

I am sure I looked on one of the calcs yesterday and saw Arabic as one of our mixes. I cannot find Arabic now so maybe I saw it in my sleep lol :P

My Gran was K1a mtDNA and hadn't thought about it until now, but if 'bottlenecked' in villages many of my lines could also have been K1a!

That might give you a little Mediterranean. Could also be you held on to a bit of that Alpine autosomal DNA to go along with your Y, which could pull you southeast.

rod
12-29-2013, 10:46 PM
Eurogenes K15 put me 100% West Scotland
1134

Dodecad V3 50% France and 50% Germany
1135

Dodecad K12b gave me 7/8 British Isles and the choice of 1/8 Sicilian, Italian, or Ashkenazi

HarappaWorld put me in Utah

Sein
12-29-2013, 11:24 PM
So, is 31.34° N, 64.93° E a good proxy for your ancestry; is 33.90 N, 67.73°E better?

I'd say both are fairly accurate.

Táltos
12-30-2013, 05:24 AM
This does pull me east. Most likely because of the unknown admixture.
http://farm4.staticflickr.com/3824/11614196666_77e216d712_o_d.jpg
MDLP World-22 Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 95.9% Serbian (derived) + 4.1% North-European-Mesolithic (ancestral) @ 1.1
2* 87.8% Montenegrin (derived) + 12.2% Finnish-North (derived) @ 1.27
3 94.8% Serbian (derived) + 5.2% Saami (derived) @ 1.27
4 83.3% CEU_V (derived) + 16.7% Jew-Iraqi (derived) @ 1.28
5 83.6% CEU_V (derived) + 16.4% Jew-Iran (derived) @ 1.29
6 95.3% Serbian (derived) + 4.7% Bra2 (derived) @ 1.29
7 76.3% German-North (derived) + 23.7% Cypriot (derived) @ 1.33
8 76.2% German-North (derived) + 23.8% Jew_Syria (derived) @ 1.37
9 63.3% Gagauz (derived) + 36.7% Swedish_V (derived) @ 1.38
10 82.7% CEU_V (derived) + 17.3% Jew_Georgia (derived) @ 1.45

Eurogenes K13 Oracle - Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 85.6% German + 14.4% Armenian @ 1.98
2 85.7% German + 14.3% Georgian_Jewish @ 2.06
3 83.1% German + 16.9% Turkish @ 2.12
4 85.3% German + 14.7% Kurdish @ 2.3
5 85.9% German + 14.1% Assyrian @ 2.37
6 85.5% German + 14.5% Azeri_Jewish @ 2.37
7 86.7% German + 13.3% Abhkasian @ 2.52
8 86.1% German + 13.9% Georgian @ 2.57
9 83.2% German + 16.8% Kumyk @ 2.62
10 85.6% German + 14.4% Iranian @ 2.71

Dodecad K12b Oracle – Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1* 69.3% Norwegian (Dodecad) + 30.7% Cypriots (Behar) @ 1.56
2 53.5% Bulgarian (Dodecad) + 46.5% Argyll (1000Genomes) @ 1.64
3 79.3% German (Dodecad) + 20.7% Sephardic_Jews (Behar) @ 1.77
4 60.9% Swedish (Dodecad) + 39.1% S_Italian_Sicilian (Dodecad) @ 2.09
5 54.6% Bulgarian (Dodecad) + 45.4% Orcadian (HGDP) @ 2.09
6 76.3% German (Dodecad) + 23.7% S_Italian_Sicilian (Dodecad) @ 2.09
7 54.5% Bulgarian (Dodecad) + 45.5% Orkney (1000Genomes) @ 2.1
8 56.4% Romanians (Behar) + 43.6% Argyll (1000Genomes) @ 2.14
9 82.3% German (Dodecad) + 17.7% Cypriots (Behar) @ 2.18
10 54.5% Bulgarian (Dodecad) + 45.5% Irish (Dodecad) @ 2.19

HarappaWorld Oracle - Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 82.7% utahn-white (1000genomes) + 17.3% georgia-jew (behar) @ 1.32
2 82.9% utahn-white (1000genomes) + 17.1% azerbaijan-jew (behar) @ 1.34
3 82.6% utahn-white (1000genomes) + 17.4% armenian (behar) @ 1.44
4 80% utahn-white (1000genomes) + 20% turk (behar) @ 1.55
5 79.7% utahn-white (1000genomes) + 20.3% turk-kayseri (hodoglugil) @ 1.59
6 81.6% utahn-white (hapmap) + 18.4% lebanese-druze (haber) @ 1.6
7* 79.2% orcadian (hgdp) + 20.8% druze (hgdp) @ 1.64
8 79.1% utahn-white (1000genomes) + 20.9% turk-istanbul (hodoglugil) @ 1.64
9 79.6% utahn-white (hapmap) + 20.4% cypriot (behar) @ 1.66
10 82.8% utahn-white (1000genomes) + 17.2% assyrian (harappa) @ 1.67
Thanks DMXX for this useful exercise. Now I get it!
"Free time" may be rare once you enter a residency program. So double thanks!
Yes I agree this was a useful exercise. I always have problems with thinking about admixture like plots on a map. Doing this myself helped me to see it. Joe I am glad the way you did your lines look similar to mine. So I feel like I didn't screw up my map too much! Your map is way cooler though. You have the little guy sitting in Africa looking at a map, and then you have the compass up in Russia! I don't really use graphics much on the computer. So I decided to use Paint and I kept forgetting to shut the eraser off. Well there I was erasing all the borders in Eastern Europe saying to myself like these borders haven't changed enough!

Joe B
12-30-2013, 05:59 AM
Yes I agree this was a useful exercise. I always have problems with thinking about admixture like plots on a map. Doing this myself helped me to see it. Joe I am glad the way you did your lines look similar to mine. So I feel like I didn't screw up my map too much! Your map is way cooler though. You have the little guy sitting in Africa looking at a map, and then you have the compass up in Russia! I don't really use graphics much on the computer. So I decided to use Paint and I kept forgetting to shut the eraser off. Well there I was erasing all the borders in Eastern Europe saying to myself like these borders haven't changed enough!
Not only do we plot to the same general area of Europe using the Oracles, our EEF-WHG-ANE ancestry proportions are pretty close too.
Táltos
EEF=55.67025
WHG=30.18771
ANE=14.14204

Joe
EEF 56.04
WHG 29.86
ANE 14.10

Pretty good mix.

Táltos
12-30-2013, 06:03 AM
Not only do we plot to the same general area of Europe using the Oracles, our EEF-WHG-ANE ancestry proportions are pretty close too.
Táltos
EEF=55.67025
WHG=30.18771
ANE=14.14204

Joe
EEF 56.04
WHG 29.86
ANE 14.10

Pretty good mix Good mix, I agree! And yes I thought we plotted similar on the map using the Oracles. :beerchug:

Tolan
01-03-2014, 05:40 AM
I used the method of DMXX for each of my chromosomes.
Two possibilities:
- I consider this map is inconsistent. I would put my chromosomes at random, I would not do better!
- I try to find consistency and I see the very ancient origin of my ancestors (yellow on the map below): a mixture of Corded and Cardial Ware cultures? Or other thing...
1155

R.Rocca
01-13-2014, 08:30 PM
Pretty neat approach. I am 5/8 Italian (mostly southern but some northern) and 3/8 Spaniard (northern). I wind up somewhere over Sardinia, but I have very large GDs with Sardinians. In all admixture runs (Eurogenes, Decodad etc.) I have the lowest GD distances with either Northern Italians or Tuscans.

http://r1b.org/imgs/Admixture.png

Stephen1986
01-14-2014, 07:41 AM
I would like to do this, but I have a Chromebook so I don't think I can download things like paint.net or similar software. If I posted some results, would someone be willing to create a map for me out of them?

vettor
01-14-2014, 07:58 AM
Pretty neat approach. I am 5/8 Italian (mostly southern but some northern) and 3/8 Spaniard (northern). I wind up somewhere over Sardinia, but I have very large GDs with Sardinians. In all admixture runs (Eurogenes, Decodad etc.) I have the lowest GD distances with either Northern Italians or Tuscans.

http://r1b.org/imgs/Admixture.png

IIRC, instructions was to use the capitals of the nations selected. ie from capital to capital and then select the percentage.
If you used the capitals you would fit better.

or this another system?

R.Rocca
01-14-2014, 02:43 PM
IIRC, instructions was to use the capitals of the nations selected. ie from capital to capital and then select the percentage.
If you used the capitals you would fit better.

or this another system?

In the case where the Gedmatch results show a country, I think that is OK. However, in my case some returned very specific regions (e.g. "Spain-Asturias", "Spain-Catalonia", "Northern Italy", etc.), so I think capitals would be less accurate. I expected to be somewhere in the middle of the Mediterranean between Italy and Spain anyway given that these types of mappings to not really handle water and islands all that well. Like I said, if you project my points, northward, they almost all land up somewhere between SE France and Tuscany, which is what I would expect given my mix.

Telfermagne
01-20-2014, 07:06 PM
The process seems a bit like a pseudo-average. It seems to assume that national capitals would harbour some kind of "native-genetic archetype" to compare oneself against (this pertains to the question, "what reason is there that I ought to choose the capital over some other city?"); and if one considers that capitals tend to be diversity pools resulting from contemporary commerce anyways, then there's an obvious kink in the chain.

What can more accurately be stated with admixture tests is that "I can infer that out of the populations made available to such and such a test, x% of my tested aDNA, +-z%, matches closer to specific people from such and such place in a given database than to specific people from other populations in that database.

Additionally it's troublesome for 23andMe results, how does one assign a capital to "nonspecific Northern Europe"? One can't. So only those who use Gedmatch or FTDNA can even do this.




You can do something similar with Population Finder purely algebraically as it were, essentially by inputting the populations (I use their capital cities) and their respective weights, then getting a GPS coordinate.

For example, my own PF:

Continent (Subcontinent) Population Percentage MOE
Europe Finnish, French, Orcadian, Romanian, Russian 40.62% ±26.27%
Europe (Southern European) Sardinian, Tuscan 59.38% ±26.27%

Then I derive lat and long and paste into Excel:

CITY LAT LONG WEIGHT
Helsinki 60.207 24.917 0.08124
Moscow 55.75 37.6167 0.08124
Paris 48.908 2.329 0.08124
Bucharest 44.418 26.082 0.08124
Lerwick 60.163 -1.1 0.08124
AVERAGE 53.8892 17.96894 0.4062 Karsin, Poland

Cagliari 39.232 9.141 0.2969
Florence 43.866 11.162 0.2969
AVERAGE 41.549 10.1515 0.5938 Tyrhennian Sea

WEIGHTED 46.56158924 13.32694413 1.000 Möderndorf, Austria

This is a reasonable average of my known ethnicities.

AJL
01-20-2014, 08:11 PM
"what reason is there that I ought to choose the capital over some other city?"

Not a kink in the chain at all. The capital is often one of the most populous cities and towards the centre of population, and so makes a good proxy. If you know what the population centroid or the actual region of origin is, by all means use it, but if you are faced with something like "Russian - Russia," I can assure you that Moscow or St. Petersburg will make a far better choice than Kamchatka.

As to your second point about diversity, you did not think this through. A city's diversity does not matter in the least because one is not concerned about the diversity of the actual city, just where you fit. If Paris is 50 times more diverse than Limoges, your results are still not going to change much because of Paris's diversity, and Paris is a much better choice than a completely arbitrary point on the map that is far from the population centre (e.g. Bayonne).

Joe B
01-20-2014, 09:34 PM
Not a kink in the chain at all. The capital is often one of the most populous cities and towards the centre of population, and so makes a good proxy. If you know what the population centroid or the actual region of origin is, by all means use it, but if you are faced with something like "Russian - Russia," I can assure you that Moscow or St. Petersburg will make a far better choice than Kamchatka.

As to your second point about diversity, you did not think this through. A city's diversity does not matter in the least because one is not concerned about the diversity of the actual city, just where you fit. If Paris is 50 times more diverse than Limoges, your results are still not going to change much because of Paris's diversity, and Paris is a much better choice than a completely arbitrary point on the map that is far from the population centre (e.g. Bayonne).
What AJL says about choosing population centers makes a lot of sense.
Population density maps are another tool that can help you make an educated guess.
This 2000 Population Density map of Turkey from NASA and Columbia University is a good example. It shows that something just south of the line between Ankara and Istanbul would be a good guess for the population center of Turkey.http://sedac.ciesin.columbia.edu/downloads/maps/grump-v1/grump-v1-population-density/turdens.jpg
A great sorce for population density maps.
NASA Socioeconomic Data and Applications Center (SEDAC) Population Density Maps
http://sedac.ciesin.columbia.edu/data/set/grump-v1-population-density/maps

Telfermagne
01-20-2014, 09:55 PM
Not a kink in the chain at all. The capital is often one of the most populous cities and towards the centre of population, and so makes a good proxy. If you know what the population centroid or the actual region of origin is, by all means use it, but if you are faced with something like "Russian - Russia," I can assure you that Moscow or St. Petersburg will make a far better choice than Kamchatka

As to your second point about diversity, you did not think this through. A city's diversity does not matter in the least because one is not concerned about the diversity of the actual city, just where you fit. If Paris is 50 times more diverse than Limoges, your results are still not going to change much because of Paris's diversity, and Paris is a much better choice than a completely arbitrary point on the map that is far from the population centre (e.g. Bayonne).

Makes sense, I had a complete brain fart there - utter brain fart....

Humanist
01-20-2014, 09:56 PM
What AJL says about choosing population centers makes a lot of sense.
Population density maps are another tool that can help you make an educated guess.
This 2000 Population Density map of Turkey from NASA and Columbia University is a good example. It shows that something just south of the line between Ankara and Istanbul would be a good guess for the population center of Turkey.

I agree that population density is something to keep in mind. However, the genetic center is what we are most concerned with. Using Iraqi Mandaeans as an example of how this (modern capitals as the genetic center) can affect one's placement. Iraqi Mandaeans have lived in the southern region of what is today Iraq for (at least) the better part of the last two millennia. Their language attests to this fact, as it is similar to (long extinct) Babylonian varieties of Aramaic. McDonald does not have Iraqi Mandaeans as a reference population, but he does have Armenians. He uses the modern Armenian capital of Yerevan as the population center. According to Dr. McDonald's population percentages, Iraqi Mandaeans are genetically similar to Armenians, and their coordinates are therefore shifted hundreds of miles to the north of their homeland.

Telfermagne
01-20-2014, 10:31 PM
I think I did this right (using PF results for now):

Coordinates are from Geohack:

Kirkwall 58.982135, -2.959418. 0.289366667
Madrid 40.383333, -3.716667. 0.289366667
Paris 48.8567, 2.3508. 0.289366667
Average 49.4073893, -4.325285. 0.8681. In the English channel.

Florence 43.783333, 11.25. 0.026380000
Helsinki 60.170833, 24.9375. 0.026380000
Moscow 55.75, 37.616667. 0.026380000
Bucharest 44.4325, 26.103889. 0.026380000
Cagliari 39.25, 9.05. 0.026380000
Average 48.6773332, 21.7916112. 0.1319. In Bánovce nad Ondavou, Slovakia.

My “genetic location”: 49.311095, -0.88046639. Mandeville-en-Bessin, Basse-Normandie, France.

Joe B
01-20-2014, 11:22 PM
I agree that population density is something to keep in mind. However, the genetic center is what we are most concerned with. Using Iraqi Mandaeans as an example of how this (modern capitals as the genetic center) can affect one's placement. Iraqi Mandaeans have lived in the southern region of what is today Iraq for (at least) the better part of the last two millennia. Their language attests to this fact, as it is similar to (long extinct) Babylonian varieties of Aramaic. McDonald does not have Iraqi Mandaeans as a reference population, but he does have Armenians. He uses the modern Armenian capital of Yerevan as the population center. According to Dr. McDonald's population percentages, Iraqi Mandaeans are genetically similar to Armenians, and their coordinates are therefore shifted hundreds of miles to the north of their homeland.No one system seems to work for everybody and sometimes it's just a matter of history. What did you come up with for an Iraqi Mandaean genetic population center? Bagdad for the modern era or something around the marshlands for a historical context?
Population density is more related to the genetic population center when nations or geographic areas are referenced. EUtest Oracle results for me will often site a country like Turkey or Armenia as a secondary population and a density map is a useful way to add weight to where that genetic population center should be. IMO

ZephyrousMandaru
01-21-2014, 01:24 AM
I agree that population density is something to keep in mind. However, the genetic center is what we are most concerned with. Using Iraqi Mandaeans as an example of how this (modern capitals as the genetic center) can affect one's placement. Iraqi Mandaeans have lived in the southern region of what is today Iraq for (at least) the better part of the last two millennia. Their language attests to this fact, as it is similar to (long extinct) Babylonian varieties of Aramaic. McDonald does not have Iraqi Mandaeans as a reference population, but he does have Armenians. He uses the modern Armenian capital of Yerevan as the population center. According to Dr. McDonald's population percentages, Iraqi Mandaeans are genetically similar to Armenians, and their coordinates are therefore shifted hundreds of miles to the north of their homeland.

What are the placements for the Iraqi Mandaeans on McDonald's BGA? The problem with McDonald's BGA is that lacks an adequate reference base, unlike the Dodecad/Eurogenes/Harappa Genetic Ancestry Projects. It seems rather anachronistic at this point, Eurogenes SPA is a far better at gauging biogeographical placements than McDonald's BGA is.

Telfermagne
01-29-2014, 03:08 AM
Tried it with the 4 Ancestors oracles for Dodecad and Eurogenes

Buggenhout, Flanders. 50.9946895, 4.20290033.

Dodecad V3 -
50% Argyll (substituting Argyll with Ireland since the population centroid page doesn't have data for Argyll) +25% North_Italian +25% Swedish @ 1.682
Eurogenes K13 -
50% Irish + 25% German + 25% Southwest_French @ 1.804

Using population centroid:
http://cs.ecs.baylor.edu/~hamerly/software/europe_population_weighted_centers.txt

Dodecad -
Ireland 53.115425619273, -7.4069761943761. 0.50
Italy 42.878891770737, 12.129225229324. 0.25
Sweden 58.888444437865, 15.535572129725. 0.25
Average 51.6275873, 6.75260703. Hünxe, North Rhine-Westphalia, Germany. Point A.

Eurogenes -
Ireland 53.115425619273, -7.4069761943761. 0.50
Germany 50.846581236184, 9.6860092676237. 0.25
France 47.123369318629, 2.6805478522224. 0.25
Average 50.361792, 1.65319364. Conchil-le-Temple, Nord-Pas-de-Calais, Northern France. Point B.

Telfermagne
02-04-2014, 03:48 AM
Tried it with Ancestry Composition, couldn't find a population centroid for West African countries so I used Lagos, Nigeria. The end result was Thury, Basse-Normandie, France.

23andMe Ancestry Composition (Speculative Threshold):
52.2% British & Irish
11.6% French & German
3.1% Scandinavian
29.1% Nonspecific Northern European
1.1% Iberian
0.5% Ashkenazi (23andMe points to Poland)
2.2% Nonspecific European
0.1% West African
>0.1% Unassigned

Using population centroid:
http://cs.ecs.baylor.edu/~hamerly/software/europe_population_weighted_centers.txt

United Kingdom 52.735567708185, -1.6789861408734. 0.261 + 0.0415714286 + 0.0022 = 0.304771429
Ireland 53.115425619273, -7.4069761943761. 0.261 + 0.0415714286 + 0.0022 = 0.304771429
France 47.123369318629, 2.6805478522224. 0.058 + 0.0415714286 + 0.0022 = 0.101771429
Germany 50.846581236184, 9.6860092676237. 0.058 + 0.0415714286 + 0.0022 = 0.101771429
Sweden 58.888444437865, 15.535572129725. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Norway 61.089862500005, 9.9201984517371. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Denmark 55.853126914967, 10.864851918353. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Spain 39.674578415577, -3.2843582298931. 0.0055 + 0.0022 = 0.0077
Portugal 39.704754462102, -9.1655868359197. 0.0055 + 0.0022 = 0.0077
Poland (Ashkenazi) 51.713706077277, 19.309306988862. 0.005 + 0.0022 = 0.0072
West Africa (no access to centroid data, I will use Lagos, Nigeria instead) 6.453056, 3.395833. 0.001
Average 47.0180426, 4.5324033. Thury, Basse-Normandie, France. Point C.

Location: Thury, Basse-Normandie, France. 47.0180426, 4.5324033.

Telfermagne
02-05-2014, 02:33 PM
Made an error in the above post, should have put Thury in Burgundy, not Normandy.

Cboatman
02-07-2014, 12:19 AM
Hi All,

I stumbled on this thread after trying to find something that can explain my GED.com admixture results.

Can anyone help?

Sorry I'm very new to all of this so I'm learning as I go.



What does this mean???

MDLP World-22 Oracle results:

Admix Results (sorted):

# Population Percent
1 Sub-Saharian 77.69
2 Atlantic_Mediterranean_Neolithic 5.26
3 North-East-European 4.78
4 Near_East 3.67
5 Pygmy 3.01
6 Indian 1.27
7 Samoedic 1.01
8 Paleo-Siberian 0.94
9 Austronesian 0.51
10 South-African 0.48
11 West-Asian 0.46
12 Mesoamerican 0.35
13 South-America_Amerind 0.33
14 Indo-Iranian 0.22

Single Population Sharing:

# Population (source) Distance
1 Lemba (derived) 14.03
2 Bantu (derived) 17.05
3 Mandenka (derived) 24.95
4 Sub-Saharian (ancestral) 25.06
5 Yoruba (derived) 25.06
6 Biaka_Pygmies (derived) 33.67
7 Ethiopian (derived) 55.1
8 Jew-Ethiopia (derived) 58.76
9 Jew_Ethiopia (derived) 59.83
10 Lumbee (derived) 61.86
11 Moroccan (derived) 65.9
12 Mozabite (derived) 71.97
13 Yemen (derived) 72.96
14 Australian (derived) 74.92
15 Egyptian (derived) 77.79
16 Puerto-Rican (derived) 78.67
17 Jordanian (derived) 83.07
18 Ste7 (derived) 83.97
19 Miwok (derived) 84.95
20 Mexican (derived) 85.96

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 87.6% Lemba (derived) + 12.4% British (derived) @ 3.77
2 87.5% Lemba (derived) + 12.5% CEU (derived) @ 3.78
3 87.6% Lemba (derived) + 12.4% Orcadian (derived) @ 3.78
4 82% Lemba (derived) + 18% Lumbee (derived) @ 3.79
5 87.7% Lemba (derived) + 12.3% Welsh (derived) @ 3.88
6 87.5% Lemba (derived) + 12.5% French (derived) @ 3.89
7 87.5% Lemba (derived) + 12.5% German-South (derived) @ 3.96
8 87.6% Lemba (derived) + 12.4% Norwegian_V (derived) @ 3.99
9 87.7% Lemba (derived) + 12.3% CEU_V (derived) @ 4.09
10 87.6% Lemba (derived) + 12.4% German_V (derived) @ 4.15
11 87.8% Lemba (derived) + 12.2% Swedish (derived) @ 4.17
12 87.6% Lemba (derived) + 12.4% Austrian (derived) @ 4.18
13 87.3% Lemba (derived) + 12.7% Swiss (derived) @ 4.18
14 88.1% Lemba (derived) + 11.9% Basque (derived) @ 4.2
15 87.8% Lemba (derived) + 12.2% German-North (derived) @ 4.27
16 87.5% Lemba (derived) + 12.5% Provancestralal (derived) @ 4.28
17 87.8% Lemba (derived) + 12.2% Swedish_V (derived) @ 4.32
18 87.6% Lemba (derived) + 12.4% Spaniard (derived) @ 4.35
19 87.7% Lemba (derived) + 12.3% German (derived) @ 4.37
20 87.6% Lemba (derived) + 12.4% Hungarian (derived) @ 4.39

Anglecynn
02-07-2014, 03:51 AM
Here's what i got with Eurogenes EUtest V2, 4 ancestors oracle. I find V1 is more accurate, and puts my family more to the west. I'll do ones with Dodecad and Eutest V1 tommorow.
Green markers are my paternal Grandfather, red markers are myself, the red marker with black centre is my average. I used the ones that were actual countries on the weighted long/lat list, as some of the others i didn't want to guesstimate.
http://i628.photobucket.com/albums/uu7/Brodir93/AncestrylocationsEUtestV2.png

Telfermagne
03-04-2014, 12:15 AM
Did it again between MDLP World-22, Dodecad V3, and Eurogenes K13

My Gedmatch kit number: M033622

Average Result
Genetic Location: 51.6759887, 12.7014061. Bad Schmiedeberg, Saxony-Anhalt, Germany.

Considering:

German_V + German_V + Kosovar + Swedish @ 1.573
Saratov 51.533333, 46.016667
Weight: 0.50
Kosovo 42.666667, 21.166667
Weight: 0.25
Stockholm 59.35, 18.066667
Weight: 0.25
Point A: 51.1833333, 28.416667. Grushevskogo, Osny, Zhitomirskaya oblast', Ukraine

North_Italian + Swedish + Argyll + Argyll @ 1.682
Helensburgh 56.012785, -4.730957
Weight: 0.50
Milan 45.466667, 9.183333
Weight: 0.25
Stockholm 59.35, 18.066667
Weight: 0.25
Point B: 53.6098173, 7.50634767. Ochtersum, Germany.

German + Irish + Irish + Southwest_French @ 1.804
Dublin 53.347778, -6.259722
Weight: 0.50
Bordeaux 44.84, -0.58
Weight: 0.25
Berlin 52.516667, 13.383333
Weight: 0.25
Point C: 50.234815, 2.18120367. Nœux-lès-Auxi, France

Telfermagne
03-26-2014, 06:08 PM
Tried this again, this time I excluded the MDLP World-22 results since they are inconsistent with my results from the other services (Even though the results between 23andMe, Dodecad, Eurognes, & Dr. McDonald vary, they have been similar overall. My MDLP results were quite discrepant.).

Result: Northern French

47.2027149, 3.02055719. La Charité-sur-Loire, Nièvre, France.
51.6275873, 6.75260703. Hünxe, Nordrhein-Westfalen, Germany.
50.361792, 1.65319364. Conchil-le-Temple, Nord-Pas-de-Calais, France.
49.9294685, 0.50078086. Butot-Vénesville, Seine-Maritime, France.
-----------------------------------------------------------------------------------------
49.7803905, 2.9817847. Voyennes, Picardy, France

Point A (based on 23andMe): 47.2027149, 3.02055719. La Charité-sur-Loire, Nièvre, France.
United Kingdom 52.735567708185, -1.6789861408734. 0.261 + 0.0415714286 + 0.0022 = 0.304771429
Ireland 53.115425619273, -7.4069761943761. 0.261 + 0.0415714286 + 0.0022 = 0.304771429
France 47.123369318629, 2.6805478522224. 0.058 + 0.0415714286 + 0.0022 = 0.101771429
Germany 50.846581236184, 9.6860092676237. 0.058 + 0.0415714286 + 0.0022 = 0.101771429
Sweden 58.888444437865, 15.535572129725. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Norway 61.089862500005, 9.9201984517371. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Denmark 55.853126914967, 10.864851918353. 0.0103333333 + 0.0415714286 + 0.0022 = 0.0541047619
Spain 39.674578415577, -3.2843582298931. 0.0055 + 0.0022 = 0.0077
Portugal 39.704754462102, -9.1655868359197. 0.0055 + 0.0022 = 0.0077
Poland (Ashkenazi) 51.713706077277, 19.309306988862. 0.005 + 0.0022 = 0.0072
Sierra Leone 8.48445, -13.23445. 0.001

52.2% British & Irish
11.6% French & German
3.1% Scandinavian
29.1% Nonspecific Northern European
1.0% Iberian
0.1% Nonspecific Southern European
0.5% Ashkenazi
2.2% Nonspecific European
0.1% West African
-----------------------------------------------------------------------------------------
47.2027149, 3.02055719. La Charité-sur-Loire, Nièvre, France.

Point B (based on my best fit from the Dodecad V3 Oracle-4): 51.6275873, 6.75260703. Hünxe, Nordrhein-Westfalen, Germany.
* Here I have chosen to use Ireland’s population centroid over that of the United Kingdom (The U.K.’s centroid is in Derbyshire, England. The folk of Argyll seem to be the Scottish analogue to the folk of County Antrim, and as such are more akin to the Irish than to the folk of the England’s East Midlands).
Ireland 53.115425619273, -7.4069761943761. 0.50
Sweden 58.888444437865, 15.535572129725. 0.25
Italy 42.878891770737, 12.129225229324. 0.25

50% Argyll +25% North_Italian +25% Swedish @ 1.682
-----------------------------------------------------------------------------------------
51.6275873, 6.75260703. Hünxe, Nordrhein-Westfalen, Germany.

Point C (based on my best fit from the Eurogenes K13 Oracle-4): 50.361792, 1.65319364. Conchil-le-Temple, Nord-Pas-de-Calais, France.
Ireland 53.115425619273, -7.4069761943761. 0.50
Germany 50.846581236184, 9.6860092676237. 0.25
France 47.123369318629, 2.6805478522224. 0.25

50% Irish +25% German +25% Southwest_French @ 1.804
-----------------------------------------------------------------------------------------
50.361792, 1.65319364. Conchil-le-Temple, Nord-Pas-de-Calais, France.

Point D (based on my best fit from Dr. McDonald’s BGA): 49.9294685, 0.50078086. Butot-Vénesville, Seine-Maritime, France.
United Kingdom 52.735567708185, -1.6789861408734. 0.50
France 47.123369318629, 2.6805478522224. 0.50

50% English + 50% French
-----------------------------------------------------------------------------------------
49.9294685, 0.50078086. Butot-Vénesville, Seine-Maritime, France.

Telfermagne
06-14-2014, 04:17 PM
Took me a while to realize that I did something wrong in the past calculations, I did not apply the weights.

Got some corrected figures (will share final result first, then scratchwork):

Final Result: 51.501312, 3.457005. Zoutelande, Veere, Zeeland, Netherlands.

Point A (derived from Dr. McDonald’s BGA): 49.83976, 1.931228. Thieulloy-l'Abbaye, Somme, Picardy, Northern France.
Point B (derived from 23andMe’s Ancestry Composition): 53.00458, 0.234492. Wrangle, Boston, Lincolnshire, England.
Point C (derived from FTDNA’s My Origins): 49.039013, 7.465819. Bitche, Moselle, Lorraine, France.
Point D (derived from Dodecad V3 Oracle-4): 54.52872, 3.990108. North Sea.
Point E (derived from Eurogenes K13 Oracle-4): 51.148692, 7.320629. Hückeswagen, Oberbergischen Kreis, North Rhine-Westphalia, Germany
Point F (derived from Eurogenes K15 Oracle-4): 51.193108, -0.43136. Guilford, Surrey, England.

The results from each utility has been given its own section, and I substituted a given admixture component with a geographic coordinate. The coordinate is a population centroid. I recycled the percentage of a given component and used it as the weight on a line between centroids, and I found a geographic midpoint between the aforementioned coordinates. I then found a geographic midpoint between the midpoints found for each section.


Point A Scratchwork: 49.83976, 1.931228. Thieulloy-l'Abbaye, Somme, Picardy, Northern France.

French= 0.468 English= 0.532 or
French= 0.701 Irish= 0.299 or
French= 0.856 Lithuanian= 0.144 or
English= 0.855 Tuscan= 0.145 or
English= 0.822 Italian= 0.178 or
Spain= 0.164 English= 0.836 or
French= 0.877 Finland= 0.123 or
French= 0.830 Belorus= 0.170 or
Irish= 0.633 Italian= 0.367

Population Centroids for Europe can be found here:
http://cs.ecs.baylor.edu/~hamerly/software/europe_population_weighted_centers.txt.

I then utilized geomidpoint to calculate the final coordinates: http://www.geomidpoint.com
*Input the centroid Latitudes & Longitudes and their Weights from each section (1-9) and let the utility work. Then input the resulting outcomes from each section (1-9) to get the final coordinates for Point A. Since baylor.edu does not have a centroid for Tuscany, I must calculate this.
Tuscany’s Centroid = 43.64881, 10.979886 Castelfiorentino, Florence, Tuscany, Italy

*Total Pop: 3,749,813
Province of Arezzo (350,000) Weight = 0.09333799
Province of Florence (933,860) Weight = 0.24904175
Province of Grosseto (227,498) Weight = 0.06066916
Province of Livorno (343,003) Weight = 0.09147203
Province of Lucca (400,000) Weight = 0.10667199
Province of Masa Carrara (200,000) Weight = 0.05333599
Province of Pisa (420,554) Weight = 0.11215333
Province of Pistoia (300,000) Weight = 0.08000399
Province of Prato (250,000) Weight = 0.06666999
Province of Siena (300,000) Weight = 0.08000399

French= 0.468 English= 0.532 or
France 47.123369318629, 2.6805478522224. Weight 0.468
United Kingdom 52.735567708185, -1.6789861408734. Weight 0.532
Weighted outcome: 50.129608, 0.488233
French= 0.701 Irish= 0.299 or
France 47.123369318629, 2.6805478522224. Weight 0.701
Republic of Ireland 53.115425619273, -7.4069761943761. Weight 0.299
Weighted outcome: 49.001476, -0.072546
French= 0.856 Lithuanian= 0.144 or
France 47.123369318629, 2.6805478522224. Weight 0.856
Lithuania 55.226946021432, 23.888296821978. Weight 0.144
Weighted Outcome: 48.498122, 5.261828
English= 0.855 Tuscan= 0.145 or
United Kingdom 52.735567708185, -1.6789861408734. Weight 0.855
Tuscany 43.64881, 10.979886. Weight 0.145
Weighted Outcome: 51.516764, 0.444627
English= 0.822 Italian= 0.178 or
United Kingdom 52.735567708185, -1.6789861408734. Weight 0.822
Italy 42.878891770737, 12.129225229324. Weight 0.178
Weighted Outcome: 51.119494, 1.175419
Spain= 0.164 English= 0.836 or
United Kingdom 52.735567708185, -1.6789861408734. Weight 0.836
Spain 39.674578415577, -3.2843582298931. Weight 0.164
Weighted Outcome: 50.605767, -1.99939
French= 0.877 Finland= 0.123 or
France 47.123369318629, 2.6805478522224. Weight 0.877
Finland 61.728144020734, 25.050425796452. Weight 0.123
Weighted Outcome: 49.080951, 4.632439
French= 0.830 Belarus= 0.170 or
France 47.123369318629, 2.6805478522224. Weight 0.830
Belarus 53.534380910258, 27.811009864462. Weight 0.170
Weighted Outcome: 48.562919, 6.421648
Irish= 0.633 Italian= 0.367
Republic of Ireland 53.115425619273, -7.4069761943761. Weight 0.633
Italy 42.878891770737, 12.129225229324. Weight 0.367
Weighted Outcome: 49.762797, 0.674308

Point B Scratchwork: 53.00458, 0.234492. Wrangle, Boston, Lincolnshire, England.
23andMe ID: Seth Reeder (Cadwallon)

52.2% British & Irish
11.6% French & German
3.1% Scandinavian
29.1% Nonspecific Northern European
-------------------------------------------------------------------
1.1% Iberian
0.5% Ashkenazi
2.2% Nonspecific European
0.1% West African

United Kingdom 52.735567708185, -1.6789861408734. Weight 0.261 + 0.0415714286 + 0.0031428571 = 0.305714286
Republic of Ireland 53.115425619273, -7.4069761943761. Weight 0.261 + 0.0415714286 + 0.0031428571 = 0.305714286
Weighted Outcome1: 52.959934, -4.530406. Weight = 0.611428572
France 47.123369318629, 2.6805478522224. Weight 0.058 + 0.0415714286 + 0.0031428571 = 0.102714286
Germany 50.846581236184, 9.6860092676237. Weight 0.058 + 0.0415714286 + 0.0031428571 = 0.102714286
Weighted Outcome1: 49.037944, 6.052218. Weight = 0.205428572
Sweden 58.888444437865, 15.535572129725. Weight 0.0103333333 + 0.0415714286 + 0.0031428571 = 0.055047619
Norway 61.089862500005, 9.9201984517371. Weight 0.0103333333 + 0.0415714286 + 0.0031428571 = 0.055047619
Denmark 55.853126914967, 10.864851918353. Weight 0.0103333333 + 0.0415714286 + 0.0031428571 = 0.055047619
Weighted Outcome1: 58.633564, 12.117512. Weight = 0.165142857
Spain 39.674578415577, -3.2843582298931. Weight 0.0055 + 0.0031428571 = 0.0086428571
Portugal 39.704754462102, -9.1655868359197. Weight 0.0055 + 0.0031428571 = 0.0086428571
Weighted Outcome1: 39.726771, -6.224329. Weight = 0.0172857142
Ashkenazi (based on Polish cities with notable Jewish populations) 51.906906, 18.550952. Weight = 0.005
West Africa 8.48445, -13.23445. Weight = 0.001
Final Weighted Outcome = 53.00458, 0.234492

Point C Scratchwork: 49.039013, 7.465819. Bitche, Moselle, Lorraine, France
FTDNA ID: 160082
100% European Coastal Plain
With my result I received a heatmap that shows the distribution of the European Coastal Plain component. The component seems to be most intense in Switzerland. There also seems to be a significant enough presence throughout the entirety of France, Germany, Switzerland, Benelux, Austria, Czech Republic, Slovenia, and Croatia to justify using the centroids for all of the mentioned countries with rough weights assigned based on the apparent intensity of the component in a given place.

Total = 100%
Switzerland 47.022583268843, 7.9530278276365. Weight = 0.25
France 47.123369318629, 2.6805478522224. Weight = 0.125
Germany 50.846581236184, 9.6860092676237. Weight = 0.125
Belgium 50.844559858929, 4.4371970671517. Weight = 0.125
Luxembourg 49.644268461894, 6.0836201926406. Weight = 0.125
Netherlands 52.069769909488, 5.2856484460503. Weight = 0.125
Austria 47.765193698288, 14.634123877631. Weight = 0.03125
Czech Republic 49.821788117433, 15.613237699363. Weight = 0.03125
Slovenia 46.167165804776, 14.886165339852. Weight = 0.03125
Croatia 45.317641918746, 16.262944879793. Weight = 0.03125
Weighted Outcome = 49.039013, 7.465819

Scratchwork Point D: 54.52872, 3.990108. North Sea
Gedmatch ID: M033622
To save time I will be using only my best fit from a given Oracle-4 run.
1 North_Italian + Swedish + Argyll + Argyll @ 1.682

I need to find a centroid N. Italy. Given the comparatively small size and isolation of Argyll, there seems to be little point in finding a centroid for it.

N. Italy Total Pop = 27,213,372
Valle d'Aosta, Italy (128,000). Weight = 0.00470357
Piedmont, Italy (4,456,000). Weight = 0.16374303
Liguria, Italy (1,617,000). Weight = 0.05941932
Lombardy, Italy (9,871,287). Weight = 0.36273664
Emilia-Romagna, Italy (4,429,000).Weight = 0.16275087
Veneto, Italy (4,936,000).Weight = 0.18138142
Friuli-Venezia Giulia, Italy (1,235,000).Weight = 0.0453821
Trentino Alto-Adige, Italy (1,037,000).Weight = 0.03810627

Argyll 56.25, -5.25. Weight = 0.50
Northern Italy 45.343983, 10.115874. Weight = 0.25
Sweden 58.888444437865, 15.535572129725. Weight = 0.25
Weighted Outcome = 54.52872, 3.990108

Scratchwork Point E: 51.148692, 7.320629. Hückeswagen, Oberbergischen Kreis, North Rhine-Westphalia, Germany

Best fit = 1 French_Basque + North_German + Norwegian + West_German @ 1.977

Given that Basque Country largely contains a population isolate there seems to be little point in finding a centroid for it. I will find a centroid for West Germany and Northern Germany.

Northern Germany Total Pop (from adding the populations of major cities only) = 8,053,181
Berlin (3,460,725). Weight = 0.42973392
Hamburg (1,786,448). Weight = 0.22183135
Bremen (547,340). Weight = 0.06796569
Hanover (522,686). Weight = 0.06490429
Bielefeld (323,270). Weight = 0.0401419
Muenster (279,803). Weight = 0.03474441
Brunswick (248,867). Weight = 0.03090294
Kiel (239,526). Weight = 0.02875249
Magdeburg (231,549). Weight = 0.02875249
Luebeck (210,232). Weight = 0.02610546
Rostock (202,735). Weight = 0.02517452

Western Germany Total Pop (from adding the populations of major cities only) = 6,623,240
Cologne (1,007,119). Weight = 0.15205836
Frankfurt am Main (697,664). Weight = 0.10533576
Duesseldorf (588,735). Weight = 0.08888927
Dortmund (580,444). Weight = 0.08763747
Essen (574,635). Weight = 0.08676041
Duisburg (489,599). Weight = 0.07392137
Bochum (374,737). Weight = 0.05657911
Bielefeld (323,270). Weight = 0.04880844
Muenster (279,803). Weight = 0.04224564
Wiesbaden (275,976). Weight = 0.04166782
Aachen (258,664). Weight = 0.039054
Moenchengladbach (257,993). Weight = 0.03895269
Gelsenkirchen (257,981). Weight = 0.03895088

Basque Country 42.881944, -1.916944. Weight = 0.25
Norway 61.089862500005, 9.9201984517371. Weight = 0.25
Northern Germany 52.861678, 11.367848. Weight = 0.25
Western Germany 51.148692, 7.320629. Weight = 0.25

Weighted Outcome = 51.148692, 7.320629

Scratchwork Point F: 51.193108, -0.43136. Guilford, Surrey, England.

Best fit = 1 Southwest_English + Southwest_English + Southwest_English + West_German @ 4.753

Since I calculated a centroid for Western Germany above, I only need to do this for Southwest England.

Southwest England Total Pop = 5,289,000
Gloucestershire (598,300). Weight = 0.11312157
Bristol (433,100). Weight = 0.08188694
Wiltshire (684,000). Weight = 0.12932501
Somerset (910,200). Weight = 0.17209302
Dorset (413,800). Weight = 0.07823785
Devon (747,900). Weight = 0.14140669
Cornwall (536,000). Weight = 0.10134241
Isles of Scilly (2,200). Weight = 0.00041596

Southwest England 51.094303, -3.005314. Weight = 0.75
Western Germany 51.148692, 7.320629. Weight = 0.25

Weighted Outcome = 51.193108, -0.43136.

AJL
06-14-2014, 04:56 PM
Hi All,

I stumbled on this thread after trying to find something that can explain my GED.com admixture results.

Can anyone help?

Is your known ancestry something like southern/central African, or are you of substantially African colonial origins?

AJL
06-14-2014, 05:03 PM
Ashkenazi (based on Polish cities with notable Jewish populations) 51.906906, 18.550952. Weight = 0.005

This will not affect your outcome a great deal, but the Ashkenazi centroid would much more accurately be placed not in Poland, but somewhere in a triangle between:

Nicosia, Cyprus +35.1667° N, +33.3667° E
Urfa/Şanlıurfa, Turkey +37.1583° N, +38.7917° E
Damascus, Syria +33.5130° N, +36.2920° E

Telfermagne
06-14-2014, 05:27 PM
Thanks for the suggestion, I've consistently been confused as to how to consider the Ashkenazi "centroid". I mimicked 23andMe's Ancestry Composition highlight & FTDNA's Jewish Diaspora heat map just to save time. I did something similar for the West African centroid as well. For the sake of speed I just picked Sierra Leon if I recall correctly (I figured that since it's a negligible value in that it occurs only once in the dataset with a value of 0.1%, a precise centroid wasn't really necessary to get a general idea for my origins). I'm of like mind in that values which are that small won't change the outcome by a significant measure, perhaps a change would shift my spot to another town in Zeeland.


This will not affect your outcome a great deal, but the Ashkenazi centroid would much more accurately be placed not in Poland, but somewhere in a triangle between:

Nicosia, Cyprus +35.1667° N, +33.3667° E
Urfa/Şanlıurfa, Turkey +37.1583° N, +38.7917° E
Damascus, Syria +33.5130° N, +36.2920° E

Asparuk
06-19-2014, 03:33 AM
I'm swimming with dolphins somewhere south of Crimea

http://i57.tinypic.com/350w7c2.png

icebreaker
06-19-2014, 05:01 AM
http://i60.tinypic.com/218ncn.jpg

Oracle results
MDLP WORLD-22 oracle
1 96.2% Turk (derived) + 3.8% Yukagir (derived) @ 2.2
Eurogenes k13
1 82.1% Turkish + 17.9% Turkmen @ 2.34
Dodecad k12b
1 72.4% Turkish (Dodecad) + 27.6% Turkmens (Yunusbayev) @ 1.79
2 85.9% Turks (Behar) + 14.1% Tajiks (Yunusbayev) @ 2.26
3 78.3% Turks (Behar) + 21.7% Turkmens (Yunusbayev) @ 2.37
4 82.4% Turkish (Dodecad) + 17.6% Tajiks (Yunusbayev) @ 2.64
5 90.7% Turks (Behar) + 9.3% Hazara (HGDP) @ 2.99
6 89.1% Turks (Behar) + 10.9% Uzbeks (Behar) @ 3.02
Harappa
1 76.5% turk-kayseri (hodoglugil) + 23.5% turkmen (yunusbayev) @ 2.07
2 85% turk-kayseri (hodoglugil) + 15% tajik (yunusbayev) @ 2.37
3 72.1% turkish (harappa) + 27.9% turk-aydin (hodoglugil) @ 2.59
4 57.6% turk-aydin (hodoglugil) + 42.4% kurd (xing) @ 2.67
5 91.5% turk-kayseri (hodoglugil) + 8.5% brahui (hgdp) @ 2.74
6 90.6% turk-kayseri (hodoglugil) + 9.4% burusho (hgdp) @ 2.75
7 91% turk-kayseri (hodoglugil) + 9% balochi (hgdp) @ 2.77
8 90.8% turk-kayseri (hodoglugil) + 9.2% makrani (hgdp) @ 2.79
9 93.4% turkish (harappa) + 6.6% kazakh (harappa) @ 2.82

My paternal line is from Trabzon (1/8), maternal line from Afyon (7/8).

MfA
07-28-2014, 01:48 PM
Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 79.7% Armenian + 20.3% Afghan_Pashtun @ 2.21
2 84% Armenian + 16% Kalash @ 2.32
3 75.8% Kurdish + 24.2% Armenian @ 2.54
4 82.1% Armenian + 17.9% Balochi @ 2.55
5 82.2% Armenian + 17.8% Brahui @ 2.66

Dodecad K12b
# Primary Population (source) Secondary Population (source) Distance
1 79% Iranian (Dodecad) + 21% Cypriots (Behar) @ 1
2 59.7% Iranians (Behar) + 40.3% Turkish (Dodecad) @ 2.13
3 80.9% Kurd (Dodecad) + 19.1% Cypriots (Behar) @ 2.2
4 82% Iranian (Dodecad) + 18% Sephardic_Jews (Behar) @ 2.31
5 55% Iranians (Behar) + 45% Turks (Behar) @ 2.39

HarappaWorld
# Primary Population (source) Secondary Population (source) Distance
1 93.3% kurd (yunusbayev) + 6.7% sardinian (hgdp) @ 1.95
2 94.1% kurd (yunusbayev) + 5.9% tunisia (henn2012) @ 2.36
3 87.2% kurd (yunusbayev) + 12.8% morocco-jew (behar) @ 2.37
4 86.2% kurd (yunusbayev) + 13.8% sephardic-jew (behar) @ 2.48
5 93.8% kurd (yunusbayev) + 6.2% morocco-n (henn2012) @ 2.5
6 78.6% iranian (harappa) + 21.4% cypriot (behar) @ 2.58
7 91.8% kurd (xing) + 8.2% sardinian (hgdp) @ 2.61
8 86.6% kurd (harappa) + 13.4% cypriot (behar) @ 2.67
9 69.1% kurd (harappa) + 30.9% armenian (harappa) @ 2.8
10 93.8% kurd (yunusbayev) + 6.2% algeria (henn2012) @ 2.8
11 55.6% iranian (harappa) + 44.4% armenian (harappa) @ 2.81
12 83.3% kurd (harappa) + 16.7% armenian (behar) @ 2.82
13 87.4% kurd (yunusbayev) + 12.6% ashkenazy-jew (behar) @ 2.86
14 94.3% kurd (yunusbayev) + 5.7% mozabite (hgdp) @ 2.9
15 95% kurd (yunusbayev) + 5% basque (hgdp) @ 2.93
16 95% kurd (yunusbayev) + 5% spain-basc (henn2012) @ 2.97
17 93.4% kurd (yunusbayev) + 6.6% puerto-rican (1000genomes) @ 2.98
18 94.4% kurd (yunusbayev) + 5.6% saharawi (henn2012) @ 2.98
19 78.6% kurd (harappa) + 21.4% turk-kayseri (hodoglugil) @ 2.99
20 83.1% kurd (xing) + 16.9% sephardic-jew (behar) @ 3.01

MDLP-22
# Primary Population (source) Secondary Population (source) Distance
1 79.2% Kurd (derived) + 20.8% Jew_Romania (derived) @ 3.12
2 79.5% Kurd (derived) + 20.5% Ashkenazim (derived) @ 3.19
3 82.4% Kurd (derived) + 17.6% Ashkenazim_V (derived) @ 3.31
4 73.2% Azeri (derived) + 26.8% Jew_Syria (derived) @ 3.34
5 79.7% Azeri (derived) + 20.3% Jew_Tunisia (derived) @ 3.39
6 80.3% Kurd (derived) + 19.7% Sicilian (derived) @ 3.41
7 81.3% Kurd (derived) + 18.7% Greek_South (derived) @ 3.44
8 80.7% Azeri (derived) + 19.3% Jew_Libya (derived) @ 3.44
9 75.2% Kurd (derived) + 24.8% Greek_Cretan (derived) @ 3.48
10 80.8% Kurd (derived) + 19.2% Italian-South (derived) @ 3.49
11 80.3% Azeri (derived) + 19.7% Jew_Morocco (derived) @ 3.52
12 79.8% Kurd (derived) + 20.2% Greek_East (derived) @ 3.55
13 79.9% Kurd (derived) + 20.1% Jew_Francestrale (derived) @ 3.56
14 81.2% Kurd (derived) + 18.8% Italian-Center (derived) @ 3.56
15 87.5% Kurd (derived) + 12.5% Portugese (derived) @ 3.57
16 79.3% Kurd (derived) + 20.7% Jew_Italia (derived) @ 3.59
17 80.6% Azeri (derived) + 19.4% Sephardim (derived) @ 3.59
18 88.1% Kurd (derived) + 11.9% Iberian (derived) @ 3.62
19 80.8% Kurd (derived) + 19.2% Greek_Center (derived) @ 3.62
20 80.5% Azeri (derived) + 19.5% Jew_Algeria (derived) @ 3.64




http://abload.de/img/minymju4.png

AJL
07-28-2014, 04:57 PM
Not the best art, but here is mine, finally. I should say that my McDonald spot is from about 2008-2009, early days, and to be fair to him, I always get a rather large margin of error.
2143

AJL
07-28-2014, 05:10 PM
Hi All,

I stumbled on this thread after trying to find something that can explain my GED.com admixture results.

Can anyone help?

Sorry, I don't think anyone ever answered you -- I am guessing you are either African American, or southern African with some European ancestry?

AJL
07-28-2014, 05:14 PM
Thanks for the suggestion, I've consistently been confused as to how to consider the Ashkenazi "centroid".

It would probably be best to avoid oracle results with Ashkenazim as DMXX originally suggested -- however Lebanese Christians (centroid a little northeast of Beirut) and Samaritans (centroid Nablus, Palestine) seem to be two of the closer populations.

vettor
07-28-2014, 06:27 PM
Not the best art, but here is mine, finally. I should say that my McDonald spot is from about 2008-2009, early days, and to be fair to him, I always get a rather large margin of error.
2143

we seem to always be nearby to each other ............my mcdonald spot was the border of France, Italy and Switzerland.

Maybe you should try Interpretome and see what difference it makes. My Interpretome places me on the border of Switzerland , Italy and Austria.
While this method in this thread puts me around Innsbruck Austria

AJL
07-28-2014, 07:51 PM
we seem to always be nearby to each other ............my mcdonald spot was the border of France, Italy and Switzerland.

Maybe you should try Interpretome and see what difference it makes. My Interpretome places me on the border of Switzerland , Italy and Austria.
While this method in this thread puts me around Innsbruck Austria

Very close to me:

http://www.anthrogenica.com/showthread.php?2289-Interpretome-POPRES-European&p=33244&viewfull=1#post33244

MitchellSince1893
07-29-2014, 03:53 AM
Here's mine. Matches that I typically come closest to Danes, Dutch, and Germans on genetic distances.

Corresponds nicely with Old Saxony and the Angles homeland i.e Anglo-Saxons

Mamluk
07-31-2014, 04:21 AM
My father's:
2165

MDLP:
2166

Harappa:
2167

Eurogenes:
2168

Dodecad:
2169

AJL
07-31-2014, 04:25 PM
^ So many Jordanian matches -- I wonder if the Jordanian sample set was in fact largely people of West Bank Palestinian descent, since this is about half the population of Jordan?

DMXX
07-31-2014, 04:54 PM
Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 79.7% Armenian + 20.3% Afghan_Pashtun @ 2.21
2 84% Armenian + 16% Kalash @ 2.32
...


A quick thought; have any other forum Iranians and Kurds attempted to model the various Iranian and Kurdish samples with the Oracle fixed to Armenian? They probably serve as a rough proxy for the indigenous West Asians.

Mamluk
07-31-2014, 10:30 PM
^ So many Jordanian matches -- I wonder if the Jordanian sample set was in fact largely people of West Bank Palestinian descent, since this is about half the population of Jordan?

That very well could be. Many Jordanians of Palestinian descent refer to themselves as "Jordanian" now, some of my cousins included. Their reasoning is that our Galilean families have been in Jordan for the past 3 generations (since 1920's British Mandate), the new generations have never even set foot in Palestine, and they are citizens of Jordan. Most West Bank families have not been in Jordan that long though (most since 1967). Amman is jokingly called "New Nablus."

But the military outpost of Upper Galilee (i.e. our ancestral town) did not have as much socio-economic interaction with Jerusalem and its surroundings as it did with Beirut and Damascus.

I think the Jordanian matches are resulting from my paternal grandfather's mother, who was a "real Arab" from the Zidan clan (http://en.wikipedia.org/wiki/Daher_el-Omar).

I am a bit baffled by so many Spanish, French/French Basque matches, among others (SW England), and (a bit disappointed) I'm not seeing traces of Balkan ancestors that my grandmother spoke so proudly of. Also interesting are the various Jewish group matches, including Ashkenazi... makes me wonder if those are from genes shared from the Holy Land, ... or perhaps from somewhere else? ;)

AJL
07-31-2014, 10:53 PM
Also interesting are the various Jewish group matches, including Ashkenazi... makes me wonder if those are from genes shared from the Holy Land, ... or perhaps from somewhere else? ;)

An open question still. :)

Faustus
08-01-2014, 08:22 PM
I hate to be "that guy", but does someone want to do the mapping for me? I'm not so great with computers. It would be greatly appreciated. Here are my results.

77.2% West_Scottish + 22.8% Spanish_Murcia (Eurogenes)
69.6% German-South (derived) + 30.4% German (derived) (MDLP)
85.5% Orkney (1000 Genomes) + 14.5% Sicilian (Dodecad)

Sapporo
08-02-2014, 05:08 AM
I tried doing this once before but failed drastically. Anyone willing to help out? :)

Mixed Mode Population Sharing:

Dodecad K12b:

# Primary Population (source) Secondary Population (source) Distance
1 63.6% Pathan (HGDP) + 36.4% Brahmins_from_Uttar_Pradesh (Metspalu) @ 1.95
2 69.2% Pathan (HGDP) + 30.8% Kshatriya (Metspalu) @ 2.9
3 78.2% Pathan (HGDP) + 21.8% Kanjars (Metspalu) @ 3.2
4 78.2% Pathan (HGDP) + 21.8% Muslim (Metspalu) @ 3.34
5 84.7% Pathan (HGDP) + 15.3% Chamar (Metspalu) @ 3.35


HarappaWorld:

# Primary Population (source) Secondary Population (source) Distance
1 58.6% pathan (hgdp) + 41.4% up-brahmin (harappa) @ 1.8
2 62.1% punjabi-jatt-sikh (harappa) + 37.9% punjabi-brahmin (harappa) @ 1.95
3 64.4% pathan (hgdp) + 35.6% brahmin-uttar-pradesh (metspalu) @ 1.97
4 66.7% punjabi-brahmin (harappa) + 33.3% pathan (hgdp) @ 2.04
5 64.9% punjabi-jatt-sikh (harappa) + 35.1% kashmiri-pandit (reich) @ 2.15

Eurogenes K13:

# Primary Population (source) Secondary Population (source) Distance
1 76.9% Kalash + 23.1% North_Kannadi @ 3.23
2 80.5% Pathan + 19.5% Gujarati @ 3.34
3 73.1% Kalash + 26.9% Kanjar @ 3.35
4 79.6% Kalash + 20.4% Chamar @ 3.43
5 87.6% Pathan + 12.4% Kanjar @ 3.45
6 82.5% Pathan + 17.5% Kshatriya @ 3.45
7 64.5% Kalash + 35.5% Kshatriya @ 3.52
8 72.6% Kalash + 27.4% Dharkar @ 3.53
9 90% Pathan + 10% North_Kannadi @ 3.57
10 76.4% Kalash + 23.6% Kol @ 3.59

Eurogenes K15:

Mixed Mode Population Sharing:

# Primary Population (source)Secondary Population (source)Distance
1 80.3% Pathan +19.7% Kshatriya @3.66
2 79% Pathan +21% Gujarati @3.66
3 86.3% Pathan +13.7% Kanjar @3.68
4 78.4% Pathan +21.6% Brahmin_UP @3.72
5 88.5% Pathan +11.5% North_Kannadi @3.75

AJL
08-02-2014, 07:36 AM
I hate to be "that guy", but does someone want to do the mapping for me? I'm not so great with computers. It would be greatly appreciated. Here are my results.

77.2% West_Scottish + 22.8% Spanish_Murcia (Eurogenes)
69.6% German-South (derived) + 30.4% German (derived) (MDLP)
85.5% Orkney (1000 Genomes) + 14.5% Sicilian (Dodecad)

The first puts you in west Wales, in the Brecon Beacons National Park; the second (somewhere) in Germany; and the third somewhere near the English Channel.

AJL
08-02-2014, 07:42 AM
I tried doing this once before but failed drastically. Anyone willing to help out? :)

Given your ancestry I'd suggest running it through Harappa as well, but I believe your spot will be close to Lahore.

I-P109
08-02-2014, 11:13 AM
I hate to be "that guy", but does someone want to do the mapping for me? I'm not so great with computers. It would be greatly appreciated. Here are my results.

77.2% West_Scottish + 22.8% Spanish_Murcia (Eurogenes)
69.6% German-South (derived) + 30.4% German (derived) (MDLP)
85.5% Orkney (1000 Genomes) + 14.5% Sicilian (Dodecad)

Here you go.
http://i.imgur.com/cSUhDzJ.png

DMXX
08-02-2014, 12:45 PM
I tried doing this once before but failed drastically. Anyone willing to help out? :)


No problem my friend. Please find the map below. I've also included which Mixed Mode results I utilised with the reasoning if necessary. Would be interested to know if this is consistent with your ancestry (pred. Punjabi with a demonstrated slight pull to the south unless I'm misremembering your results?)



Dodecad K12b:
1 63.6% Pathan (HGDP) + 36.4% Brahmins_from_Uttar_Pradesh (Metspalu) @ 1.95
HarappaWorld:
1 58.6% pathan (hgdp) + 41.4% up-brahmin (harappa) @ 1.8 <<<<<----------- Both the UP Brahmin locations were centred at Lucknow.
Eurogenes K13:
1 76.9% Kalash + 23.1% North_Kannadi @ 3.23
Eurogenes K15:
2 79% Pathan +21% Gujarati @3.66 <<------------ Have no idea where the Kshatriya sample came from, so this was a more appropriate choice.



http://i60.tinypic.com/im00te.png

Mamluk
08-02-2014, 02:33 PM
I'm not seeing traces of Balkan ancestors

I retract that statement. The high Greek percentages are probably representative of Albanian-Kosovar (Illyrian).

Humanist
08-02-2014, 03:02 PM
Geographic coordinates are not based on modern political boundaries (due to their unreliability in the case of certain populations), but rather the coordinates determined in the 2012 Eurogenes SPA run (excluding Indians and Pakistanis). Mizrahi Jews have been excluded from my coordinates, since most were not included in the Eurogenes run in question. Rather, I have used Oracle combinations that include populations like Kurds, Iranians, Cypriots, Assyrians (East and West) etc. to map my location, even though I have had to sacrifice the best fits in certain cases in order to achieve this.


MDLP
57.9% Kurd 42.1% Cypriot (3.03)


Eurogenes
79.5% Assyrian 20.5% Kurdish (1.91)



Dodecad
74.0% Assyrian 26.0% Iranian (2.71)



Harappa
86.3% Assyrian 13.7% Iranian (1.39)


http://i1096.photobucket.com/albums/g326/dok101/dok101012/humanist_point_average_map.jpg

Did a search for "Baqubah" and learned something new today!

Wikipedia


Baqubah (Aramaic: ܒܰܩܽܘܒܰܐ‎, Arabic: بعقوبة‎; BGN: Ba‘qūbah; also spelled Baquba and Baqouba, is of Syriac-Aramaic origin and means "Jacob's house" [1]) is the capital of Iraq's Diyala Governorate. The city is located some 50 km (31 mi) to the northeast of Baghdad, on the Diyala River. In 2003 it had an estimated population of some 467,900 people.[2]

dp
08-02-2014, 05:02 PM
2198
This is a rough map, based on three plot lines.
MDLP World 22: 70.3% Orcadian (derived) 29.7% Gagauz
>>Kirkwall, Orkneys (58.981, -2.96) to Chisinau, Moldova (47, 28.916667)
Eurogenes K13: 75.4% Irish, 24.6% Southwest_French
>>Dublin, Ireland (53.347778, -6.259722) to La Rochelle, France (46.1591, -1.1517)
Dodecad V3: 55.7% Swedish, 44.3% N_Italian
>>Stockholm, Sweden (59.35, 18.066667) to Milan, Italy (45.466667, 9.183333)

Did I do this right?
dp :-)

dp
08-02-2014, 05:05 PM
Did a search for "Baqubah" and learned something new today!
I thought that when Jacob spent 14 years getting his family going, he was at his father-in-law's, Laban the Syrian.

vettor
08-02-2014, 06:49 PM
mine below

http://i103.photobucket.com/albums/m153/vicpret/pllott_zps1f21c582.jpg (http://s103.photobucket.com/user/vicpret/media/pllott_zps1f21c582.jpg.html)

As previous one was was accidentally removed

I included DougM prediction for me , Interpretome and my ancestors for last 350 years (or longer)

parasar
08-02-2014, 07:33 PM
Harappa
1 65.1% gujarati (harappa) + 34.9% haryana-jatt (harappa) @ 2.23
# Population (source) Distance
1 up-brahmin (harappa) 3.43


Eurogenes k13
1 87.7% Brahmin_UP + 12.3% Brahui @ 2.54
# Population (source) Distance
1 Brahmin_UP 5.05


Dodecad k12b
1 92.1% Brahmins_from_Uttar_Pradesh (Metspalu) + 7.9% Turkmens (Yunusbayev) @ 2.23
# Population (source) Distance
1 Brahmins_from_Uttar_Pradesh (Metspalu) 4.25

All three put me to the west of UP Brahmins (closest on all calculators), while my actual location is to their east.

Sapporo
08-03-2014, 02:50 AM
@DMXX

Thank you. I really appreciate it. I don't think I really have a pull to the south. More like Central/Eastern Punjab where all my ancestry is from. Doaba/Malwa regions for the most part. It looks like I average out around Lahore/Amritsar as AJL pointed out I likely would. Also, I believe the Kshatriya is from somewhere in Uttar Pradesh.

rms2
08-03-2014, 04:16 PM
I finally got around to trying this. It was fun but some work, since I am not very skilled at using Paint.

Here it is. The red line is Oracle via MDLP. The light blue line is Oracle via Eurogenes13, and the light green line is Oracle via DodecadV3. The orange lines connects the points where my percentages fall along the lines.

2203

The result is a neat little triangle running from SW Scotland to Northern England. That is probably not too far off.

AJL
08-03-2014, 04:33 PM
^ I suspect your placement explains in large measure why my mother -- about one-third of whose ancestry is from southern Scotland, Yorkshire, North Lincolnshire, Durham, and Cumbria -- has many untraceable autosomal matches in the Carolinas and Virginia.

Humanist
08-03-2014, 05:56 PM
HarappaWorld
# Primary Population (source) Secondary Population (source) Distance
1 92.3% iranian (harappa) + 7.7% ashkenazi (harappa) @ 1.44
2 92% iranian (harappa) + 8% ashkenazy-jew (behar) @ 1.5
3 94% iranian (harappa) + 6% tuscan (1000genomes) @ 1.5
4 95% iranian (harappa) + 5% italian (hgdp) @ 1.5
5 96% iranian (harappa) + 4% spaniard (1000genomes) @ 1.54
6 94.6% iranian (harappa) + 5.4% romanian-a (behar) @ 1.6
--->7 94.5% iranian (harappa) + 5.5% bulgarian (yunusbayev) @ 1.61<---
8 96.1% iranian (harappa) + 3.9% french (hgdp) @ 1.62
9 96.8% iranian (harappa) + 3.2% spain-basc (henn2012) @ 1.69
10 96.9% iranian (harappa) + 3.1% basque (hgdp) @ 1.7

I was curious to see where you would fall if I used the SPA coordinates. The only combination that was possible to plot was the Harappa Oracle one, above.


http://i1096.photobucket.com/albums/g326/dok101/DNA_Tribes_SNP_Analysis/dmxx_.jpg

Solothurn
08-03-2014, 06:42 PM
Nice one rms2

Can you offer any easy instructions on me doing mine?

S.


I finally got around to trying this. It was fun but some work, since I am not very skilled at using Paint.

Here it is. The red line is Oracle via MDLP. The light blue line is Oracle via Eurogenes13, and the light green line is Oracle via DodecadV3. The orange lines connects the points where my percentages fall along the lines.

2203

The result is a neat little triangle running from SW Scotland to Northern England. That is probably not too far off.

Alpine Hominin
08-03-2014, 06:47 PM
I gave this a shot, and this is what I came up with:

2204

I know fairly few of my lines ancestral origins. I have ancestors from Armagh, Northern Ireland on my mothers side. My paternal line seems to be either English or Huguenot origin (My ancestors were at the Manakintowne settlement on the James River), and I have known ancestry from Bordeaux, France.

DMXX
08-03-2014, 06:56 PM
I was curious to see where you would fall if I used the SPA coordinates. The only combination that was possible to plot was the Harappa Oracle one, above.


Very interesting; I'm still placing quite a bit further east of where I "should" place based on geographical ancestry. The SPA coordinate places within the bounded zone of my "mapped location" as well.

Humanist
08-03-2014, 07:20 PM
Very interesting; I'm still placing quite a bit further east of where I "should" place based on geographical ancestry. The SPA coordinate places within the bounded zone of my "mapped location" as well.

Same here. Except I plot a few hundred miles to the south of where I would expect, based on recent geographical ancestry. But, I think both of our locations make a good deal of sense, all things considered.

vettor
08-03-2014, 07:25 PM
mine below

http://i103.photobucket.com/albums/m153/vicpret/pllott_zps1f21c582.jpg (http://s103.photobucket.com/user/vicpret/media/pllott_zps1f21c582.jpg.html)

As previous one was was accidentally removed

I included DougM prediction for me , Interpretome and my ancestors for last 350 years (or longer)

oops, I forgot to add the new YHRD heat map plotting for me........the bigger the heat the more likely

http://i103.photobucket.com/albums/m153/vicpret/smallplot_zpsf81c7da5.jpg (http://s103.photobucket.com/user/vicpret/media/smallplot_zpsf81c7da5.jpg.html)

Humanist
08-03-2014, 07:29 PM
oops, I forgot to add the new YHRD heat map plotting for me........the bigger the heat the more likely

http://i103.photobucket.com/albums/m153/vicpret/smallplot_zpsf81c7da5.jpg (http://s103.photobucket.com/user/vicpret/media/smallplot_zpsf81c7da5.jpg.html)

Do you have a direct link to this feature at YHRD?

vettor
08-04-2014, 05:56 AM
Do you have a direct link to this feature at YHRD?

I am on different time zone, thats why the delay

here is link

http://next.yhrd.org/

go to search database .

16/17 to 17/17 gives you a accurate heat map

12/17 to 15/17 gives the blue dots

the five tests are based on how many markers you tested,........... left being least and ...........right being the most

all data based of each individual, providing 5 samples each, and any fault sample in any sample will result in all samples being discarded..........so to conclude, the STR are very accurate

Tolan
08-04-2014, 08:51 AM
Thanks Vettor!
Here's my map.

Minimal: only 23 matches

2210

rms2
08-04-2014, 01:58 PM
Nice one rms2

Can you offer any easy instructions on me doing mine?

S.

I think the only way to do it is to follow DMXX's instructions step-by-step (they're in the original post at the bottom).

Really, you're just using Paint to draw lines between your top primary and secondary source populations in "Mixed Mode Population Sharing" at the bottom of your Oracle results for each run of at least three of the calculators: Eurogenes, MDLP, and Dodecad. Then you put a dot on each line to represent where you fall between the primary and secondary source populations. Finally, just connect the dots.

The hardest part for me was figuring out how to manipulate Paint.

Here are those instructions from DMXX's original post:



1. Visit GEDmatch (presuming you already have genomic raw data and have an account, otherwise, acquire these)
2. Select Admixture on the main menu.
3. Insert your ID and select any of the open-source projects (MDLP, Eurogenes, Dodecad, HarappaWorld) and then any of the calculators.
4. Once the calculator has finished, click the Oracle button (not Oracle-4) underneath your generated component scores.
5. Scroll down to the bottom and inspect the "Mixed Mode Population Sharing:" results. Pick one, preferably one with the lowest genetic difference (GD) to ensure better accuracy and one which includes non-diaspora/recently non-admixed populations (localising Ashkenazi Jewish or African Americans as a donor population on a map will be difficult due to subjective guidance regarding their placement on a map).
6. Repeat the above with at least two other calculators and keep note of the results. For a minimalist approach, Europeans are better off using Eurogenes, Dodecad and MDLP. South Asians are recommended to have HarappaWorld included. Those from elsewhere in the world are free to use any combination, as none of these are specific for other regions.
7. Download this map (http://commons.wikimedia.org/wiki/File:A_large_blank_world_map_with_oceans_marked_in _blue.PNG) (from Wikipedia) and Paint.Net (open-source image editor). Feel free to use another editing software. I prefer Paint.Net because it indicates the 1/3 increments along any line drawn.
8. Open the map with Paint.Net/another image editor. Pinpoint your McDonald BGA average spot or physical ancestral location if desired.
9. With a colour specific to the open-source calculator you're going to use, pinpoint the location where each donor population for your selected Oracle result comes from. If uncertain, look up roughly where they're from (e.g. Pakistani Pashtuns will be around NW Pakistan close to the Afghan border). If a national average (e.g. German_Dodecad), place in the middle of the country.
10. Draw a line between both donor populations. Estimate where on the line you'll fall. Note the numbers are flipped round in practice; for instance, if the Oracle is 70% German + 30% Ukrainian, the spot will end up around the 30% mark on the German end. Make a spot on the line wherever this may be.
11. Repeat steps 9+10 for all the other Oracle runs, remembering to use different colours for the calculators to keep track.
12. Join these spots together with a different coloured line, forming the "bounded area" where your ancestry can be narrowed down from.
13. Completed. Make all the relevant inferences from the results, compare to the additional data in step 8 if present.

Táltos
08-04-2014, 02:21 PM
Of course at YHRD if I try to use the 16/17 markers for my brother I get this message:
Found no match in 78,530 Haplotypes.

I'm pretty sure in the past I was able to use more markers and get an answer there. I wonder why it's like that now?

So I used just the 11 markers, and here is his map.
2211

2212

rms2
08-04-2014, 02:48 PM
I don't think yhrd ever does me much good. At so few markers my haplotype resembles the Frisian Modal Haplotype too much and I get a bunch of U106 noise, even though I am U106- (confirmed by both FTDNA and BritainsDNA).

dp
08-04-2014, 06:05 PM
Thanks guys for letting me know how to do this. I wanted to see if my map would cover different geography than my family members. So here's two more maps.
My mom's is f345900. My dad's brother is f345328. Except on mine, posted earlier, locations and percentages of the generated clines --if I'm using the term correctly-- are on the pictures.
Mom's: 2215 Mine: 2218 Uncle's: 2216
My uncle comes up basically north German. My mom's DNA includes a German component, but hers has a west-northwestward pull towards England.
I think I better understand my old FamilyFinder results. I guess when I came 91% Orcadian, 9% Middle East I was being given one cline. I wonder what population represented the Middle East?
dp :-)
PS: how do you get a single map coordinate out of the data? What are you doing, running all the lines of each admix util?

PPS:sorry about the following empty map attachment. dont see how to get rid of it.

vettor
08-04-2014, 06:59 PM
Of course at YHRD if I try to use the 16/17 markers for my brother I get this message:
Found no match in 78,530 Haplotypes.

I'm pretty sure in the past I was able to use more markers and get an answer there. I wonder why it's like that now?

So I used just the 11 markers, and here is his map.
2211

2212

did you use all the 5 tests?

at 17/17
I get 18 matches at the 1st tests ( left one )........which is 12 STR's
11 matches at next
3 matches in middle test
and zero for other 2

but heat map is based on 1st test

Humanist
08-04-2014, 08:37 PM
I am on different time zone, thats why the delay

here is link

http://next.yhrd.org/

Thank you.

Táltos
08-05-2014, 03:43 AM
did you use all the 5 tests?

I just checked them all now. I was pressed for time earlier. He only has matches with the minimal amount.

vettor
08-05-2014, 06:12 AM
I just checked them all now. I was pressed for time earlier. He only has matches with the minimal amount.

next test with , ...........ignore fast mutating STR and leave them blank..................too me , in all plotting tests regardless of what tests , I always test as per its wants, then always test withour fast mutating STRs

I get only 3 heat map areas in my maps once I ignore the fast mutating ones, they are central austria, north-east Italy and slovenia .................to me it seem logical to ignore the Fast mutating

what do you think?

rms2
08-05-2014, 01:27 PM
Thanks guys for letting me know how to do this. I wanted to see if my map would cover different geography than my family members. So here's two more maps.
My mom's is f345900. My dad's brother is f345328. Except on mine, posted earlier, locations and percentages of the generated clines --if I'm using the term correctly-- are on the pictures.
Mom's: 2215 Mine: 2218 Uncle's: 2216
My uncle comes up basically north German. My mom's DNA includes a German component, but hers has a west-northwestward pull towards England.
I think I better understand my old FamilyFinder results. I guess when I came 91% Orcadian, 9% Middle East I was being given one cline. I wonder what population represented the Middle East?
dp :-)
PS: how do you get a single map coordinate out of the data? What are you doing, running all the lines of each admix util?

PPS:sorry about the following empty map attachment. dont see how to get rid of it.

Nice work.

I'm thinking of using a nice, color political map like yours and going back and redoing mine.

Tolan
08-07-2014, 04:43 PM
I redid my map including all those which there is a oracle function in gedmatch.
I removed all the lines of construction, keeping only the points.
I took the first two results for each calculator.
There are two incoherent results compared to other (in Romania).
2266


Here my first map
http://www.anthrogenica.com/showthread.php?1778-Mapping-Your-Location-(GEDmatch-calculators)&p=24888&viewfull=1#post24888



90% of my ancestors came from Eastern Brittany (Ille et Vilaine) and 10% of the neighboring department of Mayenne.

I am obviously interested in the local history of this area (And I understand that it does not interest you! ;))

If I'm representative (by my ancestors) in this region, we can say that:
1) The bretons (from British Islands) do not have much colonized East of Britanny.
Which is quite consistent with what we know (Breton has never been spoken, the area was already populated and fairly Romanized before the Breton invasion in fourth century
2) I am more "Germanic" or "Eastern Europe" as many of my fellow citizens, while I am very western France.
Perhaps this region has a particular history:
Before 850, the area was part of the "marche of Bretagne." It was a land conflict between Bretons and the Franks, in the Frankish territory.
It is possible that there has been an influx of Frankish population (military and civilian) to keep this territory, which was, in the end ,won by Nominoë for the Bretons in 850.
Otherwise, I do not see why I would be more Eastern than the French average ..

but it is possible that I am not representative because of the vagaries of transmission of genes ..:P

Táltos
08-08-2014, 03:32 AM
next test with , ...........ignore fast mutating STR and leave them blank..................too me , in all plotting tests regardless of what tests , I always test as per its wants, then always test withour fast mutating STRs

I get only 3 heat map areas in my maps once I ignore the fast mutating ones, they are central austria, north-east Italy and slovenia .................to me it seem logical to ignore the Fast mutating

what do you think?
I didn't take out any fast mutating STRs as I figured there are already not that many to work with! I saw that you wrote in another post about there being more to come. I hope so. There are several tabs that you can't click on.

I also noticed they gave a list of SNPs and a breakdown of how many haplotypes one can find under each test. http://next.yhrd.org/tools/ysnps click on #Haplotypes
But so many haplogroups (subclades) are missing. :( Oh well they do say the aim and objectives of the website is to "generate reliable Y-STR haplotype...".

Hanna
08-08-2014, 06:05 PM
https://scontent-b-sin.xx.fbcdn.net/hphotos-xpa1/v/t1.0-9/1919645_338703786305300_6839138982187274377_n.jpg? oh=9fd69a15b6e3f286576bb890be61d8f7&oe=5463D373



Dodecad K12b

Primary Population (source) Secondary Population (source) Distance
1 62.6% Turks (Behar) + 37.4% Armenians (Behar) @ 2.54
2 83.7% Armenians (Behar) + 16.3% Uzbeks (Behar) @ 2.84
3 76.9% Armenian (Dodecad) + 23.1% Nogais (Yunusbayev) @ 2.85
4 85.7% Armenians (Behar) + 14.3% Hazara (HGDP) @ 3.17
5 58.5% Turkish (Dodecad) + 41.5% Armenians (Behar) @ 3.43
6 61.4% Turks (Behar) + 38.6% Armenian (Dodecad) @ 3.48
7 61.3% Turks (Behar) + 38.7% Armenians_15 (Yunusbayev) @ 3.5
8 77.7% Armenians_15 (Yunusbayev) + 22.3% Nogais (Yunusbayev) @ 3.54
9 55.4% Turkish (Dodecad) + 44.6% Armenians_15 (Yunusbayev) @ 3.59
10 84.6% Turks (Behar) + 15.4% Georgians (Behar) @ 3.62
11 85.9% Armenians (Behar) + 14.1% Uygur (HGDP) @ 3.66
12 75.3% Armenians (Behar) + 24.7% Turkmens (Yunusbayev) @ 3.69
13 89.5% Armenians (Behar) + 10.5% Altai (Rasmussen) @ 3.73
14 86.1% Armenian (Dodecad) + 13.9% Uzbeks (Behar) @ 3.92
15 55.9% Turkish (Dodecad) + 44.1% Armenian (Dodecad) @ 3.92
16 84% Turks (Behar) + 16% Abhkasians (Yunusbayev) @ 3.96
17 82.5% Armenians (Behar) + 17.5% Tajiks (Yunusbayev) @ 4.16
18 82.1% Turkish (Dodecad) + 17.9% Georgians (Behar) @ 4.22
19 90.1% Armenians (Behar) + 9.9% Mongol (Rasmussen) @ 4.31
20 91.1% Armenian (Dodecad) + 8.9% Altai (Rasmussen) @ 4.41

Harappa

# Primary Population (source) Secondary Population (source) Distance
1 91.8% armenian (behar) + 8.2% kazakh (harappa) @ 2.11
2 92.8% armenian (behar) + 7.2% altaian (rasmussen) @ 2.12
3 92.2% armenian (behar) + 7.8% kyrgyz (hodoglugil) @ 2.19
4 93.2% armenian (behar) + 6.8% mongolian (rasmussen) @ 2.25
5 93.6% armenian (behar) + 6.4% tuvinian (rasmussen) @ 2.28
6 93.8% armenian (behar) + 6.2% buryat (rasmussen) @ 2.3
7 69.9% turk-kayseri (hodoglugil) + 30.1% georgian (behar) @ 2.47
8 89.6% armenian (behar) + 10.4% uzbek (behar) @ 2.48
9 94.2% armenian (behar) + 5.8% oroqen (hgdp) @ 2.48
10 83.1% armenian (behar) + 16.9% nogai (yunusbayev) @ 2.51
11 91% armenian (behar) + 9% hazara (hgdp) @ 2.6
12 91.1% armenian (behar) + 8.9% uyghur (hgdp) @ 2.73
13 94.9% armenian (behar) + 5.1% yakut (hgdp) @ 2.8
14 68.2% turk-kayseri (hodoglugil) + 31.8% abhkasian (yunusbayev) @ 2.84
15 94.6% armenian (behar) + 5.4% daur (hgdp) @ 2.84
16 93.9% armenian (behar) + 6.1% yukaghir (rasmussen) @ 2.84
17 74% turk (behar) + 26% georgian (behar) @ 2.85
18 94.7% armenian (behar) + 5.3% dolgan (rasmussen) @ 2.86
19 94.6% armenian (behar) + 5.4% hezhen (hgdp) @ 2.88
20 95.3% armenian (behar) + 4.7% evenki (rasmussen) @ 2.99

Eurogenes k13 61.3% Assyrian + 38.7% Balkar @ 3.28

MDLP World 22 93.3% Turk (derived) + 6.7% Georgian_Laz (derived) @ 2.66

DMXX
08-08-2014, 06:52 PM
Dodecad K12b

Primary Population (source) Secondary Population (source) Distance
1 62.6% Turks (Behar) + 37.4% Armenians (Behar) @ 2.54
2 83.7% Armenians (Behar) + 16.3% Uzbeks (Behar) @ 2.84
3 76.9% Armenian (Dodecad) + 23.1% Nogais (Yunusbayev) @ 2.85
4 85.7% Armenians (Behar) + 14.3% Hazara (HGDP) @ 3.17
5 58.5% Turkish (Dodecad) + 41.5% Armenians (Behar) @ 3.43
6 61.4% Turks (Behar) + 38.6% Armenian (Dodecad) @ 3.48
7 61.3% Turks (Behar) + 38.7% Armenians_15 (Yunusbayev) @ 3.5
8 77.7% Armenians_15 (Yunusbayev) + 22.3% Nogais (Yunusbayev) @ 3.54
9 55.4% Turkish (Dodecad) + 44.6% Armenians_15 (Yunusbayev) @ 3.59
10 84.6% Turks (Behar) + 15.4% Georgians (Behar) @ 3.62
11 85.9% Armenians (Behar) + 14.1% Uygur (HGDP) @ 3.66
12 75.3% Armenians (Behar) + 24.7% Turkmens (Yunusbayev) @ 3.69
13 89.5% Armenians (Behar) + 10.5% Altai (Rasmussen) @ 3.73
14 86.1% Armenian (Dodecad) + 13.9% Uzbeks (Behar) @ 3.92
15 55.9% Turkish (Dodecad) + 44.1% Armenian (Dodecad) @ 3.92
16 84% Turks (Behar) + 16% Abhkasians (Yunusbayev) @ 3.96
17 82.5% Armenians (Behar) + 17.5% Tajiks (Yunusbayev) @ 4.16
18 82.1% Turkish (Dodecad) + 17.9% Georgians (Behar) @ 4.22
19 90.1% Armenians (Behar) + 9.9% Mongol (Rasmussen) @ 4.31
20 91.1% Armenian (Dodecad) + 8.9% Altai (Rasmussen) @ 4.41

Harappa

# Primary Population (source) Secondary Population (source) Distance
1 91.8% armenian (behar) + 8.2% kazakh (harappa) @ 2.11
2 92.8% armenian (behar) + 7.2% altaian (rasmussen) @ 2.12
3 92.2% armenian (behar) + 7.8% kyrgyz (hodoglugil) @ 2.19
4 93.2% armenian (behar) + 6.8% mongolian (rasmussen) @ 2.25
5 93.6% armenian (behar) + 6.4% tuvinian (rasmussen) @ 2.28
6 93.8% armenian (behar) + 6.2% buryat (rasmussen) @ 2.3
7 69.9% turk-kayseri (hodoglugil) + 30.1% georgian (behar) @ 2.47
8 89.6% armenian (behar) + 10.4% uzbek (behar) @ 2.48
9 94.2% armenian (behar) + 5.8% oroqen (hgdp) @ 2.48
10 83.1% armenian (behar) + 16.9% nogai (yunusbayev) @ 2.51
11 91% armenian (behar) + 9% hazara (hgdp) @ 2.6
12 91.1% armenian (behar) + 8.9% uyghur (hgdp) @ 2.73
13 94.9% armenian (behar) + 5.1% yakut (hgdp) @ 2.8
14 68.2% turk-kayseri (hodoglugil) + 31.8% abhkasian (yunusbayev) @ 2.84
15 94.6% armenian (behar) + 5.4% daur (hgdp) @ 2.84
16 93.9% armenian (behar) + 6.1% yukaghir (rasmussen) @ 2.84
17 74% turk (behar) + 26% georgian (behar) @ 2.85
18 94.7% armenian (behar) + 5.3% dolgan (rasmussen) @ 2.86
19 94.6% armenian (behar) + 5.4% hezhen (hgdp) @ 2.88
20 95.3% armenian (behar) + 4.7% evenki (rasmussen) @ 2.99

Eurogenes k13 61.3% Assyrian + 38.7% Balkar @ 3.28

MDLP World 22 93.3% Turk (derived) + 6.7% Georgian_Laz (derived) @ 2.66


Very similar to Imadaddin's!

Hanna
08-08-2014, 06:57 PM
Very similar to Imadaddin's!

Then why are my spots different than his?
Could you share his oracle results?

DMXX
08-08-2014, 08:13 PM
Then why are my spots different than his?
Could you share his oracle results?

I do not have possession of his Oracle results anymore, but I'm sure he'd be happy to share them whenever he is next active.

The big difference I can ascertain from your results and his is the Central Asian affinity. Imadaddin has more of it (at least 6% Uzbek from memory). The two of you have identical Oracle results in the respect that there is a predominant South Caucasian genetic base (Armenian) with significant Central Asian (Turkic speaking rather than Iranic).

This is why Imadaddin's pushed into the Caspian Sea while you're enjoying hypothetical apricots in the Armenian highlands. :P :)

rms2
08-08-2014, 09:15 PM
I tried to get a little better detail by focusing in on the intersection of all my Gedmatch results in the British Isles. This isn't an exact science (or the best artwork), but I figure it isn't too far off. It seems that autosomally I am closest to the population of SW Scotland/NW England.

2291

dp
08-08-2014, 09:38 PM
Like Tolan I decided to add some more plot lines. This altered the southern most area substantially. My plot now dips from Berlin down to central Austria, and from there to Dublin, Ireland. By doing so Belgium and a substantial portion of Germany were added. Thus my chart now looks more like a composite of my mothers and my uncles. :) Also, I noticed that 2 of the additional plot lines lay generally on the generated bound line between Austria and Dublin, so either could be removed and would not affect the outcome of the plot. To keep the map simple, I drew the clines, the indicated boundary lines, and any ending cline point that wasn't a major city I wrote into the map. The legend is the database credits and the predicted coordinate from each.

http://www.anthrogenica.com/attachment.php?attachmentid=2290&d=1407531220

I kept a datasheet as I generated the coordinates, and tried to show all my post-GEDmatch Oracle steps so others can see how I generated the boundary data points. Then I averaged the coordinates to find a maximum hot spot of my aDNA predicted ancestry. It is near the Netherlands, Belgium, German border at Dusseldorf, Germany.
To display the steps in columns I coded it into HTML. I thought that this site supported html coding when encapsulated within HTML tags, but it didn't work. You can copy the code into your text editor, save it as a .html file, and then load it with your browser. I've omitted this sites HTML tags that don't work.

<HTML><HEAD><TITLE>GEDmatch Oracle Mixed Mode Population Sharing</TITLE>
<META http-equiv=Content-Type content="text/html; charset=windows-1252">
<META content="MSHTML 5.50.4207.2601" name=GENERATOR></HEAD>
<BODY text=#000080>
<FONT size = 4>David Powell - FN87877</FONT><BR>
<TABLE cols= 7 border=1><TBODY>
<TR><TD><B>Dataset</B></TD><TD><B>Admix Output</B></TD><TD><B>Population End Coordinates</B></TD><TD><B>End Coordinates Difference (E.D.C.)</B></TD><TD><B>Majority % * E.D.C.<BR><FONT size=1>(=adjustment amount to minority population coordinates)</FONT></B></TD><TD><B>Calculate boundary coordinates<BR><FONT size=1>(in N hemisphere subtract for southwards or westwards)</FONT></B></TD><TD><B>Resulting Coordinates</B></TD></TR>

<TR><TD>Eurogenes K13</TD><TD>75.4% Irish + 24.6% Southwest_French @ 1.99</TD><TD>53.35, -6.26 + 46.16, -1.15</TD><TD>7.19, 5.11</TD><TD>5.42, 3.85</TD><TD>46.16+5.42, -1.15-3.85</TD><TD>51.56, -5</TD></TR>

<TR><TD>HarappaWorld</TD><TD>85.1% British (1000genomes)+14.9% Tuscan (1000 genomes) @ 1.17</TD><TD>51.51, -.13 + 43.78, 11.25</TD><TD>7.73, 11.38</TD><TD>6.58, 9.68</TD><TD>43.78+6.58, 11.25-9.68</TD><TD>50.36, 1.57</TD></TR>

<TR><TD>MDLP K=12</TD><TD>71.6% GER (German)+28.4% SPN (Spaniard) @ 3.33</TD><TD>52.52, 13.38 + 40.4, -3.7</TD><TD>12.12, 17.08</TD><TD>8.68, 12.23</TD><TD>40.4+8.68, -3.7+12.23</TD><TD>49.08, 8.53</TD></TR>

<TR><TD>MDLP World</TD><TD>85.4% Austrian+14.6% Provancal @ .93</TD><TD>48.2, 16.35 + 43.83, 5.78</TD><TD>4.37, 10.57</TD><TD>3.73, 9.03</TD><TD>43.83+3.73, 5.78+9.03</TD><TD>47.56, 14.81</TD></TR>

<TR><TD>Dodecad V3</TD><TD>55.7% Swedish (Dodecad)+44.3% N_Italian (Dodecad) @ .53</TD><TD>59.35, 18.07 + 45.47, 9.18</TD><TD>13.88, 8.89</TD><TD>7.73, 4.95</TD><TD>45.47+7.73, 9.18+4.95</TD><TD>53.2, 14.13</TD></TR>

<TR><TD>MDLP World-22</TD><TD>70.3% Orcadian (derived) + 29.7% Gagauz (derived) @ 1.05</TD><TD>58.98, -2.96 + 47, 28.92</TD><TD>11.98, 31.88</TD><TD>8.42, 22.41</TD><TD>47+8.42, 28.92-22.41</TD><TD>55.42, 6.51</TD></TR>

<TR><TD></TD><TD></TD><TD></TD><TD></TD><TD></TD><TD align=right><B>Average result</B></TD><TD>51.2, 6.76</B></TD></TR>

</TBODY></TABLE>
</BODY>
</HTML>

rms2
08-09-2014, 12:43 PM
Wow! Very nicely done! I may try that.

dp
08-09-2014, 04:40 PM
Richard,
a very tight geographical area :-)

DMXX
08-09-2014, 10:42 PM
Another spin on the original "Mapping Your Location" idea which (somewhat) addresses the main issue with the technique, which is the placement of each donor population's location on the empty map. This is largely at the discretion of the map-maker.

Here, I've charted my position in the exact same manner as originally described, with the difference coming from the map itself. Rather than an empty world chart, I have formatted one of Dr. McDonald's BGA charts (please visit OP to use, look for McDonald BGA version). This technique swaps real-world physical dimensions for a generic PC1xPC2 autosomal world plot. The only disadvantage here is the limited population choices (appear to be Behar et al. and HGDP only). Thus, anyone willing to try this newer version out will have to restrict their choices further.

I attempted to standardise the colour scheme, but the good doctor's previous choices made this impossible (e.g. distinction between Turks and Georgians was completely lost).

My results, shown in the spoilers, are fully in line with McDonald's prior assessment and some of the Oracle results (predominantly Iranian + <10% various European).



http://i57.tinypic.com/344y368.png



MDLP World-22
4 83.6% Turk (derived) + 16.4% Burusho (derived) @ 4.29
Secondary pops in 1-3 all Tajik (44.5-46.4%), Yunusbayev dataset, not included in BGA

Eurogenes K13
5 92.8% Iranian + 7.2% Orcadian @ 3.08
Secondary pops in 1-4 all Central, West or North Euro (7.3-8.9%), not in HGDP dataset

Dodecad K12b
6 95% Iranian (Dodecad) + 5% Russian_B (Behar) @ 3.13
Ditto MDLP results

HarappaWorld
6 94.6% iranian (harappa) + 5.4% romanian-a (behar) @ 1.6
Same reasoning as previous points

ZephyrousMandaru
08-10-2014, 07:00 AM
I was curious to see where you would fall if I used the SPA coordinates. The only combination that was possible to plot was the Harappa Oracle one, above.


http://i1096.photobucket.com/albums/g326/dok101/DNA_Tribes_SNP_Analysis/dmxx_.jpg

Interesting, would you mind mapping mine?

davidram
09-07-2014, 05:28 PM
Interesting project you guys are running here despite me not wholly getting it myself, my population distances are pretty far with most starting at a distance of 2+. David Powell suggested I post some of my runs here for a second opinion so I'm following that line of advice to see what more can be uncovered about my own background. Additionally, mtdna haplogroups are H3(h3h2) and y-chrom is e1b1a7a.

255325542555

TIA, and hope you all continue to have fun with this :)

Telfermagne
09-23-2014, 04:55 PM
Re-ran mine again, this time I took into account something that Dr. McDonald mentioned over at FTDNA forums:

"Now if the average spot on the map of a purely European/Jewish/Armenian(etc.)/Turkish person differs by more than 1000 miles between programs, then you have a really bad and exceptional case. But a 600 mile error is common and to be expected" (McDonald 2014).

I took liberty and mapped a circle around my "average location" with a 600 mile error-radius. As can be seen, it takes up quite a large chunk of Europe.

I-P109
09-23-2014, 05:07 PM
mine below

http://i103.photobucket.com/albums/m153/vicpret/pllott_zps1f21c582.jpg (http://s103.photobucket.com/user/vicpret/media/pllott_zps1f21c582.jpg.html)

As previous one was was accidentally removed

I included DougM prediction for me , Interpretome and my ancestors for last 350 years (or longer)

vettor, how did you determine the point for Interpretome? They don't seem to plot a point on a map anywhere I've seen. I'd be interested to plot mine as well.

vettor
09-23-2014, 05:57 PM
vettor, how did you determine the point for Interpretome? They don't seem to plot a point on a map anywhere I've seen. I'd be interested to plot mine as well.

apart from the "triangle" the Doug and interpretome are from other people or sites ( interpretome) and the ancestors are from my BDMs I have to ~1650

You can use Interpretome via there site, but IIRC they use 23andme and one other as reference data...................Konrad from interpretome is very helpful if you need to ask a question.

I-P109
09-23-2014, 06:23 PM
apart from the "triangle" the Doug and interpretome are from other people or sites ( interpretome) and the ancestors are from my BDMs I have to ~1650

You can use Interpretome via there site, but IIRC they use 23andme and one other as reference data...................Konrad from interpretome is very helpful if you need to ask a question.

Thanks, but what I meant was where on the Interpretome site can you obtain a data interpretation represented on a map? None of the options I've tried on the Interpretome site results in a map plotting.

vettor
09-23-2014, 06:48 PM
Thanks, but what I meant was where on the Interpretome site can you obtain a data interpretation represented on a map? None of the options I've tried on the Interpretome site results in a map plotting.

if you are european, you need to set it to european ( top right corner area ), then use PCA1 and PCA4 , select the max test and plot your marker. zoom in to your results, copy and then overlay the result on a map ........it will give you a area within 100Km .

cross reference this result with Pca1 and other PCA numbers . As other combination of numbers run in different directions......best to read the information in regards to interpretome on this site.
IIRC someone here, has a database and also has a different plotting system

I-P109
09-23-2014, 07:01 PM
if you are european, you need to set it to european ( top right corner area ), then use PCA1 and PCA4 , select the max test and plot your marker. zoom in to your results, copy and then overlay the result on a map ........it will give you a area within 100Km .

cross reference this result with Pca1 and other PCA numbers . As other combination of numbers run in different directions......best to read the information in regards to interpretome on this site.
IIRC someone here, has a database and also has a different plotting system

Sorry, but I completely lost you with the bold part. How do you copy, or rather what do you copy? There is no option to copy anything, and no data to play with. Just a plotting on a non-geographical two axis. How do you turn the two axis plot into a geographical map point?

I-P109
09-23-2014, 07:20 PM
I've read the topic here: http://www.anthrogenica.com/showthread.php?2289-Interpretome-POPRES-European&highlight=interpretome, but it does not bring any valid methodology for converting PCA plottings into geographical map ones.

vettor
09-24-2014, 06:39 AM
I've read the topic here: http://www.anthrogenica.com/showthread.php?2289-Interpretome-POPRES-European&highlight=interpretome, but it does not bring any valid methodology for converting PCA plottings into geographical map ones.


I overlapped all my PCA mixtured plots ( I think by memory I had 6 plotted which i copied ) on a plain map and that's how I got my spot.

you can even try this below...a bit of work, but fun to see the pltting trend of your ancestors
http://originhunters.com/content/MRM-BGM.pdf

hypoE
01-09-2015, 06:41 PM
How should I gauge the accuracy of doing the triangulation method? I came up with a long sliver along the Netherlands/German border reaching all the way up into Denmark. Problem being this does not match my mtdna or genealogical record. Did I perhaps use the wrong calculator settings? Harappa4, Dodecad V3, Jtest.

3346

AJL
01-09-2015, 06:53 PM
How should I gauge the accuracy of doing the triangulation method? I came up with a long sliver along the Netherlands/German border reaching all the way up into Denmark. Problem being this does not match my mtdna or genealogical record. Did I perhaps use the wrong calculator settings? Harappa4, Dodecad V3, Jtest.

3346

Are you of colonial US origins? If so that's potentially normal placement. Many colonial Americans have multiple fractions of things like Dutch, German, Italian, French, Swiss, Polish, Scandinavian, or Finnish ancestry that are enough to tug the average into that area -- though one Native American ancestor about 4-5 generations ago could have a similar effect too.

hypoE
01-09-2015, 07:15 PM
I wouldn't classify myself as colonial american, No. Predominantly French Canadian.
My second attempt using ethnihelix, MDLP World 22, and Dodecad 7b turned out what I believe to be more accurate. Centered in Southwest France, Sardinia and Northwest Italy.

3348

Salkin
01-09-2015, 07:46 PM
MDLP K23b gave me 50% Swedish + 50% Swedish as the only 2-way, so this got a bit odd. Still, it isn't completely crazy - I'm from a bit further south and west than the Swedish point I picked, but even so.

3347

vettor
01-09-2015, 08:24 PM
i tested my son's markers using a different method
I used Mdlp k23b and dodecad k12b
For each, i plotted the first 4 markers and polycon them

The results below shows where the (pink colour) Mdlp and the ( blue colour ) dodecad overlap and this corresponds nicely with his 23andme numbers

http://i103.photobucket.com/albums/m153/vicpret/hapue_zps4b356631.jpg (http://s103.photobucket.com/user/vicpret/media/hapue_zps4b356631.jpg.html)

edit same method for my wife..............to see if it overlaps with our son's above

http://i103.photobucket.com/albums/m153/vicpret/mariza_zpsddd3be41.jpg (http://s103.photobucket.com/user/vicpret/media/mariza_zpsddd3be41.jpg.html)

AJL
01-09-2015, 08:51 PM
I wouldn't classify myself as colonial american, No. Predominantly French Canadian.

That still makes sense. In essence, think of yourself as something like 13/16ths from Caen (49.44 N 1.10 E), 2/16ths from Mi'kmaq or Cree lands (a wide area to choose from but why not try Quebec City), and 1/16th from Ireland or Scotland, you should arrive at a point more or less in the area you have. This is probably a good estimation for the ancestral proportions of most typical Métis people.

Kurd
03-08-2015, 06:13 PM
My placement is around S. Khorasan Iran using the gedmatch calculators' mixed population oracles.


http://i1377.photobucket.com/albums/ah42/dil88888/Gedmatch%20location_zpsclcx6f17.jpg (http://s1377.photobucket.com/user/dil88888/media/Gedmatch%20location_zpsclcx6f17.jpg.html)

EDIT: Placement is more accurately Yazd /Khorasan provinces.

Kaido
03-08-2015, 08:12 PM
Quite a cool visualisation

http://i.imgur.com/unFJfEH.png?1

MLDP-22
90.5% Pathan (derived) + 9.5% Roma (derived) @ 2.4
eurogenes
93.5% Pathan + 6.5% Abhkasian @ 2.62
Harappa
77.8% kashmiri-pandit (reich) + 22.2% lezgin (behar) @ 1.44
dodecad
79.8% Jatt (Dodecad) + 20.2% Abhkasians (Yunusbayev) @ 4.0

Dr_McNinja
03-08-2015, 08:53 PM
Mine gave me a triangle around Haryana, India.
Eurogenes K13

# Primary Population (source) Secondary Population (source) Distance
1 81% Brahmin_UP + 19% Makrani @ 1.89
2 61.4% Brahmin_UP + 38.6% Sindhi @ 1.91
3 81.1% Brahmin_UP + 18.9% Brahui @ 1.91
4 81% Brahmin_UP + 19% Balochi @ 2.1
5 74.7% Kshatriya + 25.3% Afghan_Pashtun @ 2.18I didn't use the Baloch because the line wouldn't route through migration paths (to the northwest, then to the southwest), it'd cut straight across the subcontinent. #5 was still at a close genetic distance.


Harappa

1 84.2% punjabi-jatt-sikh (harappa) + 15.8% irula (xing) @ 1.72


Eurogenes K15

1 85.2% Kshatriya + 14.8% Chechen @ 2.24

For Uttar Pradesh I used a point halfway between New Delhi and Lucknow. For Afghan Pashtun I used a point halfway between Kandahar and Kabul. For Irula, I used Nilgiri mountainous area in South India. For Punjabi Jatt Sikh, I used Lahore.

It's hard to eyeball the distances. A script or web app which did it for you seems doable. Don't have the time to make one though :\

DMXX
03-08-2015, 09:55 PM
My placement is around S. Khorasan Iran using the gedmatch calculators' mixed population oracles.


EDIT: Placement is more accurately Yazd /Khorasan provinces.

That's an unusually eastern placement for someone of Kurdish heritage. I noticed you tend to score more in the components found more commonly in South Asia than Iranians and Kurds tend to. Are you aware of any ancestry from around the Indus in your background (noticed the flags)?

jesus
03-08-2015, 10:12 PM
Based on Harappa and Dodecad k12 I was roughly placed between Semnan, Yazd and Isfahan.
http://i.imgur.com/94wNjbu.png

Kurd
03-08-2015, 10:50 PM
That's an unusually eastern placement for someone of Kurdish heritage. I noticed you tend to score more in the components found more commonly in South Asia than Iranians and Kurds tend to. Are you aware of any ancestry from around the Indus in your background (noticed the flags)?

I tend to score similar to Jesus on some of the S Asian specific components, which is a little higher than the genotyped kurds. I would like to see greater sampling in the southern Kurdi areas though. I am not aware of any non Kurd recent ancestors either. I was a little puzzled by My DNA Tribes results also, which showed higher scores than a large percentage of Yemeni, Horn of Africa, and S Indian populations. Without anyone to compare to on that test except for a sample African American and Caucasian, they are of limited value.

The Pakistani flag signifies the birthplace of me and my wife, and where we were raised.

surbakhunWeesste
03-08-2015, 10:52 PM
I tend to score similar to Jesus on some of the S Asian specific components, which is a little higher than the genotyped kurds. I would like to see greater sampling in the southern Kurdi areas though. I am not aware of any non Kurd recent ancestors either. I was a little puzzled by My DNA Tribes results also, which showed higher scores than a large percentage of Yemeni, Horn of Africa, and S Indian populations. Without anyone to compare to on that test except for a sample African American and Caucasian, they are of limited value.

The Pakistani flag signifies the birthplace of me and my wife, and where we were raised.

Whoa! you were born in Pakistan? and you are of kurdish ancestry right! interesting :)

CelticGerman
03-09-2015, 12:12 AM
3985

GEDmatch calculators etc.

vettor
03-09-2015, 02:37 AM
3985

GEDmatch calculators etc.

what do refer to when you say Gedmatch calculators?

Leeroy Jenkins
03-09-2015, 03:23 AM
http://i60.tinypic.com/238dfs.png

Eurogenes K15: 93.4% W. Norwegian 6.6% Sardinian
Dodecad K12b: 95.6% Dutch 4.4 North Ossetian
MDLP K23b: 50% Dutch 50% Frisian

Agamemnon
03-09-2015, 03:41 AM
http://pichoster.net/images/2015/03/09/triangulation%20original%20map.jpg

The fits are:

Dodecad K12b - 59.5% Orkney (1000Genomes) + 40.5% Cypriots (Behar) @ 1.64
MDLP K23b - 50% Cretan +50% Norwegian_East @ 1.088383
Eurogenes K13 (1) - 56.4% West_Sicilian + 43.6% Danish @ 1.78
Eurogenes K13 (2) - 50.1% Southeast_English + 49.9% East_Sicilian @ 2.13

I might try mapping my mother's location as well, I'll try to do it tomorrow.

Kurd
03-09-2015, 04:23 AM
Whoa! you were born in Pakistan? and you are of kurdish ancestry right! interesting :)

I am not the first nor the last Kurd to be born there. Kurds have been in the Pakistan/ Afghanistan area for centuries. Ali Mardan Khan, a Kurd from the Rika tribe was the governor of Kandhar in the 17th century during the Safavid period. He was later appointed as governor of Kashmir, Lahore, and Kabul by Shahjahan.

Historically, Kurds have entered Pakistan through Balochistan.

Here is an excerpt of an article that I am writing on the Kurds of Balochistan (hopefully I will finish it when I have time one of these days):

According to accepted tradition, the 6 principal tribes mentioned arrived in Sarhad in the following order:
I. First, the Kurds;
II. Second, the Bamaris;
III. Third, the Daminis;
IV. Fourth, the Naruis;
V. Fifth, the Rekis;
VI. Sixth, the Ismailzais.

The Serhadi Kurd tribe consists of 3 sections:
1. Mir Balozai;
2. Sohrabzai;
3. Ghulam Shahzai.

Sardar Karim Khan belonged to the Ghulam Shahzai section. He was son of Sardar Sayad Khan, resident of Bandar Kamal Khan, on the Helmand river.
In the late 1500s, Shah Abbas sent 7 brothers from Kurdistan to Sarhad to serve as Nazims (Governors). The Kurds were ideal candidates for displacement to Baluchistan, because of their nomadic lifestyle, and their reputation for bravery and fighting. They were sent to govern the lawless areas of Baluchistan, and collect taxes for the Shah.

This made them in effect, the De Facto arm of the Persian government Baluchistan. With the help of the Governor of Bampur, they gained possession of Khash, and the neighboring areas, where they took up residence.

The fate of the 7 brothers was as follows:
1. Mir Afzal, the youngest, remained in Khash;
2. Kamar Ali, went to Kalat (Pakistan), where he joined the Brahui tribe. A subsection of that tribe, known as Kamar Alis, are his descendants;
3. Another brother went to Nushki, which is near Quetta Pakistan, and his descendant was Sardar Rasool Khan;
4. Another brother went to Kandahar, Afghanistan, and eventually took up residence in the Bolan. His descendant was Sardar Yar Mohammad, son of Sardar Allah Dinah, Kurd-Brahui;
5. The fate of the remaining 3 brothers is unknown.
Sardar Karim Khan’s mother was the elder daughter of Sardar Sharif Khan Narui of Bandar Kamal Khan.
Evidence shows that the Kurds were a dominant class in Sarhad which organized other groups under them in a complex regional system, their source of power derived from conquest of agricultural center, marital alliances with other dominant groups, and delegated authority from the state [5].
In 1809 Pottinger recorded the Kurds as the dominant group in Sarhad, and in 1838 Haji Abdun Nabi wrote; “The chief of Sarhad is Maddat Khan Kurd, originally from Persia. He collects 1/5 of the produce from the area.”[5]
Jane Bestor [5] who was a guest of the Kurd sardar of Sangan Baluchistan, Haji Amir Khan for 7 months, writes that the haji stated to her that the founding father of the tribe was Husain Kurd, who originally likely belonged to the Yelhani tribe of the Kirmanshah region of Kurdistan. Haji goes on to say that Husain, the warrior hero of folktales, went to India and fought against Emperor Akbar for 7 years, returning a hero to the court of Shah Abbas in Isfahan. Later Shah Abbas sent Husain and some Kurd families to Sarhad to claim it for Iran, and to collect tribute for the Shah.
Throughout the latter half of the 18th century, the Khanate of Kalat was the dominant presence in all of Baluchistan, and its importance was reflected in the development of Kurd alliances with its satellites, such as the Rigi and Dehani tribes, that bordered the eastern frontier of the Sarhad.
During Nadir Shah’s reign, the Hakem Mirza is said to have recruited a military force of Kurds and various border tribes, including Rigis, to fight in the Shah’s eastern campaign. As a result of the friendship which developed between the Rigi and Kurd chiefs, Mirza Khan gave permission for the Kurds to make marital alliances (wasilat) with Rigis. Kurd tradition maintains that Rigis often fought with Kurds against other groups.
In addition to the Rigis, marital ties were established with another eastern border tribe under the influence of the Khan of Kalat.

Kurd
03-09-2015, 04:29 AM
Whoa! you were born in Pakistan? and you are of kurdish ancestry right! interesting :)

Here is another excerpt from my article:

The following may have been some additional sources of post 16th century Kurd migrations to Baluchistan:
1. Descendants of tribesmen who accompanied Lotf Ali Khan Zand on his desperate flight to Bam in 1794;
2. Khorasani Kurds. According to the Cultural & Civil Society of Khorasani Kurds [3], about 2 million Kurds inhabit Khorasan.

Lotf Ali Khan Zand (1789-1794) was one of Karim Khan Zand’s (1751-1779) successors. Karim Khan was a Kurd, and the founder of the Zand dynasty. Until the 1880s, they were dominant in Khash (midway between Zahidan and Iranshahr), and their leader was known as the Sardar of the Sarhad (Sykes, pp. 106, 107, 131; see also Bestor). Today, they are widely scattered, some of them living on the southern slopes of the Kuh-e Taftan, others dwelling around Magas (today, Zabol); and still others are settled in Sistan (Afsar Sistani, p. 918). Hosayn Ali Razmara mentions eight villages in the district of Bamposht (SE of Zahidan) that are inhabited by Baluchi speaking Zand tribesmen (VIII, pp. 187, 248, 313, 315, 322, 372, 384). These probably moved to Baluchistan at the same time as the Kurds of Khash.

Afshar
03-09-2015, 10:55 AM
3986
Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 66.7% Georgian_Jewish + 33.3% Nogay @ 2.91
Dodecad K12b
# Primary Population (source) Secondary Population (source) Distance
1 52.7% Cypriots (Behar) + 47.3% Turkmens (Yunusbayev) @ 3.16
HarappaWorld
# Primary Population (source) Secondary Population (source) Distance
1 90.8% turk-istanbul (hodoglugil) + 9.2% qatari (henn2012) @ 2.3
MDLP World-22
1 96.6% Turk (derived) + 3.4% Near_East (ancestral)

Didnt know i was a Dadas all that time:)

Stephen1986
03-09-2015, 01:46 PM
My ancestry can be summarised as mostly North Western (Lancashire, Cumbria, Merseyside, Cheshire and Greater Manchester) English with significant contributions from northern and eastern Ireland (some Irish lines have unknown origins within Ireland), Gloucestershire, Suffolk, North Yorkshire and Hertfordshire, with more minor ancestry from western Wales. Two of my maternal great great grandfathers are unknown whilst there is a possibility of Scottish ancestry (I have a Scottish surname in my tree via Ireland and my maternal grandfather believed that he had some Scottish ancestry). Genetic analyses, especially for my brother, indicate some significant Continental ancestry.

Here's my map using Eurogenes K13 (shown in red), MDLP World-22 (shown in green) and Dodecad K12b (shown in blue).

For my K13 results, it showed 99%+ West Scottish plus a small amount of East African (and one had a small amount of French Basque), and the distances were all very close together (between 1.79 and 1.81). I decided to use the FB one for the sake of simplicity.

For my MDLP results, my closest result was 54.9% CEU and 45.1% Hungarian (derived), so I put CEU in the middle of the North Sea and the Hungarian in the middle of Hungary (the second one was mostly British with significant Mixed Germanic, it might have been better to use that one instead, although I don't think the triangle would look remarkably different).

For my Dodecad results, my closest match was 88.1% Cornwall (1000Genomes) and 11.9% Polish (Dodecad) with a distance of 0.29, so I put the Cornish dot in the middle of Cornwall and the Polish dot in the middle of Poland.

The triangle looks like the results my brother and myself both got from Dr MacDonald, which placed our ancestry around the area of SE England/Benelux/western Germany. It also shows a north west pull via my Irish, Northern English and possible Scottish ancestors, whilst the south western pull is likely from my Welsh and Gloucestershire ancestors.

3987

Here's my brother's map, using the same analyses and colours as my own.

For his K13 results, most of his results were similar to mine in that they were c95-99% West Scottish with small amounts of Middle Eastern populations (my brother has some ME in MyOrigins, and he shows minute traces of North African and East Asian on 23andMe). The distances ranged from 1.59 to 1.9. I decided to use the 17th result in a list of 20 which had a distance of 1.88, as I think the ME populations, like my own East African populations in the same test, are really something else that lies somewhere on the continent. This result was 55.1% Danish and 44.9% Southwest English. I put the Danish dot in the middle of Denmark and the SW English dot on Bristol.

For his MDLP results, his top five results were all mostly Norwegian or Swedish and Balkan, his closest one (with a distance of 1.49), was 50.3% Norwegian_V (derived) and 49.7% Serbian (derived). I put the Norwegian dot on Oslo and the Serbian dot on Belgrade.

For his Dodecad results, his results were 86%+ British or Orcadian and various Middle Eastern/West Asian groups. The closest match, with a distance of 0.64%, was 92.6% British_Isles (Dodecad) and 7.4% Iranians (Behar). I put the British dot in the middle of the island and the Iranian dot in the western part which is visible on the map.

It was hard to do anything more than a very narrow triangle for my brother, it looks like something is pulling him eastwards but otherwise he looks typically Northern European.

3991

Hanna
03-09-2015, 02:29 PM
@ Kurd

So this explains why you score higher Balooch and South Asian than the other Kurds. From how many generations is your family in Pakistan?

CelticGerman
03-09-2015, 06:09 PM
what do refer to when you say Gedmatch calculators?

Sorry, I don't understand your question. When I say "GEDmatch calculators" it's evident I refer to GEDmatch calculators (it's the terminology used by many people I think).

vettor
03-09-2015, 06:20 PM
Sorry, I don't understand your question. When I say "GEDmatch calculators" it's evident I refer to GEDmatch calculators (it's the terminology used by many people I think).

I meant, which ones did you use?

CelticGerman
03-09-2015, 06:36 PM
I meant, which ones did you use?

Ok, I have used EUROGENES K13 and K15, Dodecad V3 and K12b, MDLP World and World22, Harappa. In addition I have added results (genetic averages) from McDonald and David W. I tried to find the midpoint of 12 top populations listed by DNATribes (SNP) too and found out the result was quite good. I have also added the weighted midpoint according to regional percentages of my known ancestry (it's more east than the genetic midpoints; probably due to the fact many of my known Mecklenburg ancestors should have ancestors from the west, regions such as Westfalia, Netherlands, Flanders, Frisia ...).

vettor
03-09-2015, 06:40 PM
Ok, I have used EUROGENES K13 and K15, Dodecad V3 and K12b, MDLP World and World22, Harappa. In addition I have added results (genetic averages) from McDonald and David W. I tried to find the midpoint of 12 top populations listed by DNATribes (SNP) too and found out the result was quite good. I have also added the weighted midpoint according to regional percentages of my known ancestry (it's more east than the genetic midpoints; probably due to the fact many of my known Mecklenburg ancestors should have ancestors from the west, regions such as Westfalia, Netherlands, Flanders, Frisia ...).

Thanks

on DNATribes............I too have found over time that they should be taken more seriously

CelticGerman
03-09-2015, 06:44 PM
Thanks

on DNATribes............I too have found over time that they should be taken more seriously

Not their STR test (it's off), but there is real progess concerning the SNP test.

Kurd
03-09-2015, 08:37 PM
I am not the first nor the last Kurd to be born there. Kurds have been in the Pakistan/ Afghanistan area for centuries. Ali Mardan Khan, a Kurd from the Rika tribe was the governor of Kandhar in the 17th century during the Safavid period. He was later appointed as governor of Kashmir, Lahore, and Kabul by Shahjahan.

Historically, Kurds have entered Pakistan through Balochistan.

Here is an excerpt of an article that I am writing on the Kurds of Balochistan (hopefully I will finish it when I have time one of these days):

According to accepted tradition, the 6 principal tribes mentioned arrived in Sarhad in the following order:
I. First, the Kurds;
II. Second, the Bamaris;
III. Third, the Daminis;
IV. Fourth, the Naruis;
V. Fifth, the Rekis;
VI. Sixth, the Ismailzais.

The Serhadi Kurd tribe consists of 3 sections:
1. Mir Balozai;
2. Sohrabzai;
3. Ghulam Shahzai.

Sardar Karim Khan belonged to the Ghulam Shahzai section. He was son of Sardar Sayad Khan, resident of Bandar Kamal Khan, on the Helmand river.
In the late 1500s, Shah Abbas sent 7 brothers from Kurdistan to Sarhad to serve as Nazims (Governors). The Kurds were ideal candidates for displacement to Baluchistan, because of their nomadic lifestyle, and their reputation for bravery and fighting. They were sent to govern the lawless areas of Baluchistan, and collect taxes for the Shah.

This made them in effect, the De Facto arm of the Persian government Baluchistan. With the help of the Governor of Bampur, they gained possession of Khash, and the neighboring areas, where they took up residence.

The fate of the 7 brothers was as follows:
1. Mir Afzal, the youngest, remained in Khash;
2. Kamar Ali, went to Kalat (Pakistan), where he joined the Brahui tribe. A subsection of that tribe, known as Kamar Alis, are his descendants;
3. Another brother went to Nushki, which is near Quetta Pakistan, and his descendant was Sardar Rasool Khan;
4. Another brother went to Kandahar, Afghanistan, and eventually took up residence in the Bolan. His descendant was Sardar Yar Mohammad, son of Sardar Allah Dinah, Kurd-Brahui;
5. The fate of the remaining 3 brothers is unknown.
Sardar Karim Khan’s mother was the elder daughter of Sardar Sharif Khan Narui of Bandar Kamal Khan.
Evidence shows that the Kurds were a dominant class in Sarhad which organized other groups under them in a complex regional system, their source of power derived from conquest of agricultural center, marital alliances with other dominant groups, and delegated authority from the state [5].
In 1809 Pottinger recorded the Kurds as the dominant group in Sarhad, and in 1838 Haji Abdun Nabi wrote; “The chief of Sarhad is Maddat Khan Kurd, originally from Persia. He collects 1/5 of the produce from the area.”[5]
Jane Bestor [5] who was a guest of the Kurd sardar of Sangan Baluchistan, Haji Amir Khan for 7 months, writes that the haji stated to her that the founding father of the tribe was Husain Kurd, who originally likely belonged to the Yelhani tribe of the Kirmanshah region of Kurdistan. Haji goes on to say that Husain, the warrior hero of folktales, went to India and fought against Emperor Akbar for 7 years, returning a hero to the court of Shah Abbas in Isfahan. Later Shah Abbas sent Husain and some Kurd families to Sarhad to claim it for Iran, and to collect tribute for the Shah.
Throughout the latter half of the 18th century, the Khanate of Kalat was the dominant presence in all of Baluchistan, and its importance was reflected in the development of Kurd alliances with its satellites, such as the Rigi and Dehani tribes, that bordered the eastern frontier of the Sarhad.
During Nadir Shah’s reign, the Hakem Mirza is said to have recruited a military force of Kurds and various border tribes, including Rigis, to fight in the Shah’s eastern campaign. As a result of the friendship which developed between the Rigi and Kurd chiefs, Mirza Khan gave permission for the Kurds to make marital alliances (wasilat) with Rigis. Kurd tradition maintains that Rigis often fought with Kurds against other groups.
In addition to the Rigis, marital ties were established with another eastern border tribe under the influence of the Khan of Kalat.



For further information on the Kurds of Balochistan, and on the Kurdi origin of some Baloch tribes, here are some references:


1. Story of the Baloch, Dr. Naela Quadri, (http://www.cpp.net.pk/2013/06/24/story-of-the-baloch-by-professor-naela-quadri-baloch/).
2. Encyclopaedia Iranica, http://www.iranicaonline.org/articles/zand-dynasty
3. Cultural & Civil Society of Khorasani Kurds, http://cskk.org/en/modules.php?name=News&file=article&sid=7
4. Kurdish Tribes, Pierre Oberling, http://www.caissoas.com/CAIS/Anthropology/Kurds/kurdish_tribes.htm
5. Jane Bestor, The Kurds of Iranian Baluchistan: A Regional Elite, M.A. Thesis, McGill University, 1979, http://digitool.library.mcgill.ca/webclient/StreamGate?folder_id=0&dvs=1419777102201~764
6. Gazeteer of Persia Volume 4, Office of the Superintendent of Government Printing, Calcutta, 1892. https://books.google.com/books?id=QNEoAAAAYAAJ&pg=PA251&lpg=PA251&dq=kurds+of+serhad+balochistan&source=bl&ots=uGuhCqZ95F&sig=crXnKV4kvPr_QHJyVyqeHK2o2Hs&hl=en&sa=X&ei=NBugVNnjBsjxoASqoYKIAg&ved=0CDAQ6AEwBw#v=onepage&q=kurds%20of%20serhad%20balochistan&f=false;
7. History of the Baloch, http://pikalbaloch.wordpress.com/about/

dp
03-09-2015, 09:23 PM
...
It's hard to eyeball the distances. A script or web app which did it for you seems doable. Don't have the time to make one though :\
I've reformatted an earlier post of mine which may help you not have to eyeball your coordinates. -dp


Like Tolan I decided to add some more plot lines. This altered the southern most area substantially. My plot now dips from Berlin down to central Austria, and from there to Dublin, Ireland. By doing so Belgium and a substantial portion of Germany were added. Thus my chart now looks more like a composite of my mothers and my uncles. :) Also, I noticed that 2 of the additional plot lines lay generally on the generated bound line between Austria and Dublin, so either could be removed and would not affect the outcome of the plot. To keep the map simple, I drew the clines, the indicated boundary lines, and any ending cline point that wasn't a major city I wrote into the map. The legend is the database credits and the predicted coordinate from each.

http://www.anthrogenica.com/attachment.php?attachmentid=2290&d=1407531220

I kept a datasheet as I generated the coordinates, and tried to show all my post-GEDmatch Oracle steps so others can see how I generated the boundary data points. Then I averaged the coordinates to find a maximum hot spot of my aDNA predicted ancestry. It is near the Netherlands, Belgium, German border at Dusseldorf, Germany.
http://www.anthrogenica.com/attachment.php?attachmentid=3998&d=1425935932

Dr McNinja,
It's been a while since I did this. I hope it didn't have any mistakes. Hope this helps.
dp :-)

Kurd
03-09-2015, 11:24 PM
@ Kurd

So this explains why you score higher Balooch and South Asian than the other Kurds. From how many generations is your family in Pakistan?

Hanna,

My parents were born in Iraqi Kurdistan. My Baloch component is in line with some of the Southern or SE Kurds. S. Indian is a little higher, but keep in mind that I am not aware of any large scale sampling of Kurds in the most southern areas of Iraqi or Irani Kurdistan yet. It is something I look forward to.

EDIT: There is at least 1 calculator, Dodecad V3, that shows Kurd as the closest genetically (it probably uses a good set of S. Kurd references). Otherwise, Iranian is usually in 1st position. Pashtun and Tajik are down the list probably in positions >7. Makrani, Brahui, and Balochi are in positions >6.

Afshar
03-10-2015, 08:18 AM
3986
Eurogenes K13
# Primary Population (source) Secondary Population (source) Distance
1 66.7% Georgian_Jewish + 33.3% Nogay @ 2.91
Dodecad K12b
# Primary Population (source) Secondary Population (source) Distance
1 52.7% Cypriots (Behar) + 47.3% Turkmens (Yunusbayev) @ 3.16
HarappaWorld
# Primary Population (source) Secondary Population (source) Distance
1 90.8% turk-istanbul (hodoglugil) + 9.2% qatari (henn2012) @ 2.3
MDLP World-22
1 96.6% Turk (derived) + 3.4% Near_East (ancestral)

Didnt know i was a Dadas all that time:)
Its funny to see that adding more calculators results in dots on the same lines, but some (MLDP World) are a bit off and give me a point located on the Kazakhstan Caspian.

Agamemnon
03-10-2015, 06:59 PM
Here's my mother's map:

http://pichoster.net/images/2015/03/10/triangulation%20original.jpg

I used the following fits:

Eurogenes K13 - 59.7% Southwest_English + 40.3% West_Scottish @ 2.13
Eurogenes K15 - 91.8% Orcadian + 8.2% French_Basque @ 2.11
Eurogenes EUtest - 87.2% English + 12.8% IE @ 1.37
Dodecad K12b - 84.1% British_Isles (Dodecad) + 15.9% Bulgarians (Yunusbayev) @ 1.85
MDLP K23b - 50% Frisian +50% South_German @ 1.333866
MDLP World-22 - 69.3% Swedish (derived) + 30.7% Italian_North (derived) @ 0.95

CelticGerman
03-10-2015, 08:25 PM
Here is the map of my uncle maternal side. For comparison I add a map showing regions where known ancestors came from (midpoint weighted according to percentages).40004001

sweuro
03-10-2015, 09:31 PM
Here's my mother's map:

http://pichoster.net/images/2015/03/10/triangulation%20original.jpg

I used the following fits:

Eurogenes K13 - 59.7% Southwest_English + 40.3% West_Scottish @ 2.13
Eurogenes K15 - 91.8% Orcadian + 8.2% French_Basque @ 2.11
Eurogenes EUtest - 87.2% English + 12.8% IE @ 1.37
Dodecad K12b - 73.7% British_Isles (Dodecad) + 26.3% Hungarians (Behar) @ 1.7
MDLP K23b - 50% Frisian +50% South_German @ 1.333866
MDLP World-22 - 69.3% Swedish (derived) + 30.7% Italian_North (derived) @ 0.95
I think you made a mistake on K12b, that's Bulgaria not Hungary.

Agamemnon
03-10-2015, 10:19 PM
I think you made a mistake on K12b, that's Bulgaria not Hungary.

Yep, I pasted the wrong fit in my post, I'll correct that ASAP.

fil
03-11-2015, 04:48 AM
Hopefully I did it right

4004

Compared to

4005



HarappaWorld
1 77.9% velama (metspalu) + 22.1% iranian (harappa) @ 2.46
Eurogenes K13
1 80% Velamas + 20% Kurdish @ 3.25
Dodecad K12B
1 77.7% Velamas (Metspalu) + 22.3% Kurd (Dodecad) @ 2.52
Eurogenes EUTest V2 K15 (Not sure if it is the best calculator for this but MDLP give's me unusable results)
2 60.8% Piramalai + 39.2% Brahui @ 3.89

James85
04-25-2015, 09:35 AM
Can someone clarify something for me from the initial post.

Scroll down to the bottom and inspect the "Mixed Mode Population Sharing:" results. Pick one, preferably one with the lowest genetic difference (GD) to ensure better accuracy and one which includes non-diaspora/recently non-admixed populations (localising Ashkenazi Jewish or African Americans as a donor population on a map will be difficult due to subjective guidance regarding their placement on a map).

# Primary Population (source) Secondary Population (source) Distance
1 77.6% Cornish + 22.4% AT @ 1.34
2 86.5% Cornish + 13.5% Serbian @ 1.42

That's just one of my populations. Another has me at 97% German & 2.4% Kalash with a distance of 0.96.

What population am I supposed to use to map? Both? The German ... the Kalash.


Given my paper trail through marriage, etc., - I mean my father's mother's side is thoroughly Irish with a family clan history going back 1,000 years - some of the gedmatch results are either laughable or jaw dropping or as I know through both another site genetically & paper trail I am related to those whom were part of the Tsars' family eye rolling (as gedmatch says I have not a lick of obvious Russian DNA).... I'd like to know how to map my lineage via gedmatch and see if that makes anymore sense than the % results I've being getting.

James85
04-30-2015, 09:35 AM
I figured it out and am utterly puzzled

Myself - I am European of the region I expected given my history - France, Italian / Spain \ Greece and German - Ukraine regions.

My sister on the other hand has only one "map" with that area. Denmark to Spain. The others go much further off including all the way to Pakistan.



I thought gedmatch did autosomal DNA and as such "sex" shouldn't necessarily be linked ... meaning that I should have shown some connection there as well and except for a spotting of distant Jewish I do not.

Boeck
07-19-2015, 01:36 PM
So my understanding of this deal, to sum it up, you take the countries in the mixed results in Oracle. Say for example yours is German 75% and Greek 25%. Do you draw a line between Germany and Greece on the map and put a dot 1/4 from Germany to Greece?

I made an example using my mom of how I thought this is done, please correct me if there is something I should do differently

Red line: Dodecad v3 91% Argyll (1000 Genomes) + 9% Lebanese (Behar) @ 1.15

Yellow line: Eurogenes k13 73.3% Irish + 26.7% Romanian @ 2.33

Blue line: Dodecad k12 68.2% Argyll (1000Genomes) + 31.8% Bulgarians (Yunusbayev) @ 1.28

http://i619.photobucket.com/albums/tt279/german24/Mom_zps01sldu3a.png

DMXX
07-19-2015, 02:11 PM
I figured it out and am utterly puzzled

Myself - I am European of the region I expected given my history - France, Italian / Spain \ Greece and German - Ukraine regions.

My sister on the other hand has only one "map" with that area. Denmark to Spain. The others go much further off including all the way to Pakistan.

I thought gedmatch did autosomal DNA and as such "sex" shouldn't necessarily be linked ... meaning that I should have shown some connection there as well and except for a spotting of distant Jewish I do not.

The GEDmatch Oracles (or indeed any other Oracle) basically play around with the best combinations which approximate your ancestry based on the ADMIXTURE component scores. So, to make things "fit", the calculators sometimes pick off very remote populations to complete the rest of the scores you're getting.

As you correctly pointed out, literal interpretation of the results paints a contradictory picture. There's no conceivable way your biological sister has distant Pakistani ancestry, whereas you don't. The above explains why she's getting Pakistani populations shown.

The "location mapping" exercise shown here is simply a visual means of making some of these funky pairings more informative. You and your sister should both have high overlap in the "bounded area" generated.


So my understanding of this deal, to sum it up, you take the countries in the mixed results in Oracle. Say for example yours is German 75% and Greek 25%. Do you draw a line between Germany and Greece on the map and put a dot 1/4 from Germany to Greece?

I made an example using my mom of how I thought this is done, please correct me if there is something I should do differently


Perfect. Got it in one. I normally use "zoomed in" maps since it's easier to guesstimate the position, but I doubt the overall position will change to any remarkable degree.

Good job, and welcome to the forum!

Boeck
07-20-2015, 02:31 AM
The GEDmatch Oracles (or indeed any other Oracle) basically play around with the best combinations which approximate your ancestry based on the ADMIXTURE component scores. So, to make things "fit", the calculators sometimes pick off very remote populations to complete the rest of the scores you're getting.

As you correctly pointed out, literal interpretation of the results paints a contradictory picture. There's no conceivable way your biological sister has distant Pakistani ancestry, whereas you don't. The above explains why she's getting Pakistani populations shown.

The "location mapping" exercise shown here is simply a visual means of making some of these funky pairings more informative. You and your sister should both have high overlap in the "bounded area" generated.



Perfect. Got it in one. I normally use "zoomed in" maps since it's easier to guesstimate the position, but I doubt the overall position will change to any remarkable degree.

Good job, and welcome to the forum!

Were you able to notice the black dots I put on each line? They are hard to see. I couldn't figure out how to save it in the zoomed in form

dp
07-20-2015, 06:41 PM
Were you able to notice the black dots I put on each line? They are hard to see. I couldn't figure out how to save it in the zoomed in form
Based on the Oracle 2 runs you posted, you could crop the map to include only Europe. I used MS Paint to crop mine, as well as to manually add information specific to the data & results.
Yours,
dp :-)
PS: my map; related post

DMXX
07-20-2015, 06:54 PM
Were you able to notice the black dots I put on each line? They are hard to see. I couldn't figure out how to save it in the zoomed in form

Yes, which is why I cropped the original image using Paint.NET, makes mapping things a lot easier as well.

I could see the black dots just fine though (all my computer staring these days finally served a purpose I suppose).

Boeck
07-24-2015, 12:49 AM
I also made one for my aunt (mother's sister) using the same directions.

Red line: Dodecad v3 12 62.2% British_Isles (Dodecad) + 37.8% Romanians_14 (Behar) @ 1.77

Yellow line: Eurogenes k13 8 75% Irish + 25% Romanian @ 2.58

Blue line: Dodecad k12 1 84.1% British_Isles (Dodecad) + 15.9% Turks (Behar) @ 1.05



http://i619.photobucket.com/albums/tt279/german24/pat_zpsvd0fegb6.png

Boeck
07-29-2015, 02:16 AM
And this one is my uncle (their brother)

Red line: Dodecad v3 5 87.6% Orkney (1000 Genomes) + 12.4% Greek (Dodecad) @ 1.09

Yellow line: Eurogenes k13 20 59% West_German + 41% Norwegian @ 2.38

Blue line: Dodecad k12 5 80% Dutch (Dodecad) + 20% Romanians (Behar) @ 1.52

http://i619.photobucket.com/albums/tt279/german24/Tom_zpsnlqdxufl.png

Boeck
08-01-2015, 03:06 AM
And here is the mother of the three (my grandma)


Red line: Dodecad v3 11 90.7% Orkney (1000 Genomes) + 9.3% S_Italian (Dodecad) @ 1.28

Yellow line: Eurogenes k13 8 90.5% West_Scottish + 9.5% Georgian @ 2.36

Blue line: Dodecad k12 7 89.3% English (Dodecad) + 10.7% Turks (Behar) @ 0.86

http://i619.photobucket.com/albums/tt279/german24/Grandma_zpsf7ocm2yt.png

Based on this, where would their father's ancestry most likely be?

Caspian
10-22-2015, 06:24 PM
I've tried on geomidpoint using useful gedmatch calculators.
http://www.geomidpoint.com/

I've used these cities for these populations.

Azerbaijani: Tabriz (https://en.wikipedia.org/wiki/Tabriz), Iran
Iranian: Tehran (https://en.wikipedia.org/wiki/Tehran), Iran
Kurd: Sanandaj (https://en.wikipedia.org/wiki/Sanandij), Iran
Kurd_North: Amed (https://en.wikipedia.org/wiki/Diyarbak%C4%B1r), Turkey
Armenian: Yerevan (https://en.wikipedia.org/wiki/Yerevan), Armenia
Assyrian: Kirkuk (https://en.wikipedia.org/wiki/Kirkuk), Iraq
Assyrian_Arzni: Arzni (https://en.wikipedia.org/wiki/Arzni), Armenia
Lebanese (Christian, Muslim and Druze): Beirut (https://en.wikipedia.org/wiki/Beirut), Lebanon
Iraqi_Jew: Baghdad (https://en.wikipedia.org/wiki/Baghdad), Iraq
Iraqi_Mandean: Baghdad (https://en.wikipedia.org/wiki/Baghdad), Iraq
Cypriot: Nicosia (https://en.wikipedia.org/wiki/Nicosia), Cyprus
Georgian_Jew: Kutaisi (https://en.wikipedia.org/wiki/Kutaisi), Georgia
Laz_Georgia: Batum (https://en.wikipedia.org/wiki/Batumi), Georgia
Abkhasian: Sukhumi (https://en.wikipedia.org/wiki/Sukhumi), Georgia
Ossetian: Vladikavkaz (https://en.wikipedia.org/wiki/Vladikavkaz), North Ossetia
Nogai: Astrakhan (https://en.wikipedia.org/wiki/Astrakhan), Russia
Uzbek: Bukhara (https://en.wikipedia.org/wiki/Bukhara), Uzbekistan
Hazara: Bamyan (https://en.wikipedia.org/wiki/Bamyan), Afghanistan
Turkmen: Mary Region (https://en.wikipedia.org/wiki/Mary_Region), Turkmenistan
Turkmen_Uzbekistan: Khiva (https://en.wikipedia.org/wiki/Khiva), Uzbekistan

Results.
http://oi59.tinypic.com/5fglte.jpg

All locations on a map.
http://oi57.tinypic.com/e67set.jpg
http://oi58.tinypic.com/28v4l85.jpg

These results clearly show south and south-west coasts of the Caspian Sea. This region includes a part of historical Azerbaijan (Atropatene) (https://en.wikipedia.org/wiki/Atropatene) and Tabaristan (https://en.wikipedia.org/wiki/Tabaristan) also includes two important geographical factors; Mughan steppe and Alborz mountains. (https://en.wikipedia.org/wiki/Alborz)

Mainly spoken languages of this region are as follows.
Azerbaijani (Altaic? / Turkic / Oghuz / Western Oghuz)
https://en.wikipedia.org/wiki/Azerbaijani_language
Tati, Talysh, Gilaki, Mazandarani, Old Azeri (extinct) and Deilami (extinct) (Indo-European / Indo-Iranian / Iranian / Western Iranian / Northwestern Iranian / Caspian Languages)
https://en.wikipedia.org/wiki/Caspian_languages
Persian (Indo-European / Indo-Iranian / Iranian / Western Iranian / Southwestern Iranian)
https://en.wikipedia.org/wiki/Persian_language

Also this region was homeland of two important ancient people; Caspians (https://en.wikipedia.org/wiki/Caspians) and Cadusians (https://en.wikipedia.org/wiki/Cadusii).

DMXX
10-23-2015, 10:03 PM
Excellent work. I've adopted the same sample locations and calculators as yourself, with the addition of:

Adygei - Maykop, Adyghei republic, Russia
Pashtun_Afghan - Kandahar, Afghanistan
Uzbek_Afghan - Mazar-e-Sharif, Afghanistan
Brahui - Quetta, Pakistan
Tatar_Crimean - Sevastopol, Crimea
Turk - Ankara, Turkey
Assyrian - Niniveh, Iraq
Tajik - Dushanbe, Tajikistan
Tajik_Pamiri - Nau, Tajikistan
Makrani - Rumra, Pakistan





MDLP K13 Ultimate
1 Azerbaijani + Lebanese + North_Ossetian + Pashtun-Afghan @ 1.034991

MDLP K23b
1 Georgian_Jew + Georgian_Jew + Iranian + Uzbek_Afghan @ 2.499159

MDLP World 22
1 Brahui + Lebanese + Tatar_Crim + Turk @ 2.602559

MDLP World
1 Jew_Georgia + Tatar_Crim + Turk + Brahui @ 1.771243

Eurogenes K13
1 Afghan_Pashtun + Azeri + Iranian + Lebanese_Druze @ 2.491446

Eurogenes K15
1 Afghan_Pashtun + Iranian + Kurdish + Lebanese_Druze @ 3.368770

Gedrosia K11
1 Adygei + Lebanese + Makrani + Turkish @ 3.704922

Gedrosia K12
1 Iranian + Iraqi_Arab_Baghdad + Iraqi_Jew + Tajik_Pomiri @ 2.795564

Dodecad V3
1 Assyrian + Brahui + Turks + Turks @ 2.382086

Dodecad K12b
1 Georgia_Jews + Iranians + Tajiks + Uzbekistan_Jews @ 1.793558

Dodecad World 9
1 Assyrian + Iranian + Kurd + Tajiks @ 1.142684

HarappaWorld
1 assyrian + kurd + kurd + tajik @ 1.525492



My map. Although a bit more spread out than yours, note the greatest pin density's in the exact same area as yourself.

Looks like Azeris are basically just plain ol' Median-Oghuz hybrids, eh? :)

http://i.imgur.com/qle6APX.png

vettor
10-23-2015, 11:24 PM
Excellent work. I've adopted the same sample locations and calculators as yourself, with the addition of:

Adygei - Maykop, Adyghei republic, Russia
Pashtun_Afghan - Kandahar, Afghanistan
Uzbek_Afghan - Mazar-e-Sharif, Afghanistan
Brahui - Quetta, Pakistan
Tatar_Crimean - Sevastopol, Crimea
Turk - Ankara, Turkey
Assyrian - Niniveh, Iraq
Tajik - Dushanbe, Tajikistan
Tajik_Pamiri - Nau, Tajikistan
Makrani - Rumra, Pakistan





MDLP K13 Ultimate
1 Azerbaijani + Lebanese + North_Ossetian + Pashtun-Afghan @ 1.034991

MDLP K23b
1 Georgian_Jew + Georgian_Jew + Iranian + Uzbek_Afghan @ 2.499159

MDLP World 22
1 Brahui + Lebanese + Tatar_Crim + Turk @ 2.602559

MDLP World
1 Jew_Georgia + Tatar_Crim + Turk + Brahui @ 1.771243

Eurogenes K13
1 Afghan_Pashtun + Azeri + Iranian + Lebanese_Druze @ 2.491446

Eurogenes K15
1 Afghan_Pashtun + Iranian + Kurdish + Lebanese_Druze @ 3.368770

Gedrosia K11
1 Adygei + Lebanese + Makrani + Turkish @ 3.704922

Gedrosia K12
1 Iranian + Iraqi_Arab_Baghdad + Iraqi_Jew + Tajik_Pomiri @ 2.795564

Dodecad V3
1 Assyrian + Brahui + Turks + Turks @ 2.382086

Dodecad K12b
1 Georgia_Jews + Iranians + Tajiks + Uzbekistan_Jews @ 1.793558

Dodecad World 9
1 Assyrian + Iranian + Kurd + Tajiks @ 1.142684

HarappaWorld
1 assyrian + kurd + kurd + tajik @ 1.525492



My map. Although a bit more spread out than yours, note the greatest pin density's in the exact same area as yourself.

Looks like Azeris are basically just plain ol' Median-Oghuz hybrids, eh? :)

http://i.imgur.com/qle6APX.png

for fun, i copied your gedmatch selection for myself and ended up with this below

http://i103.photobucket.com/albums/m153/vicpret/my%20mid%20point_zps5k6hcu8d.jpg (http://s103.photobucket.com/user/vicpret/media/my%20mid%20point_zps5k6hcu8d.jpg.html)

it points to 1km south of https://en.wikipedia.org/wiki/Bassano_del_Grappa

Caspian
10-24-2015, 06:54 PM
Excellent work. I've adopted the same sample locations and calculators as yourself, with the addition of:

Adygei - Maykop, Adyghei republic, Russia
Pashtun_Afghan - Kandahar, Afghanistan
Uzbek_Afghan - Mazar-e-Sharif, Afghanistan
Brahui - Quetta, Pakistan
Tatar_Crimean - Sevastopol, Crimea
Turk - Ankara, Turkey
Assyrian - Niniveh, Iraq
Tajik - Dushanbe, Tajikistan
Tajik_Pamiri - Nau, Tajikistan
Makrani - Rumra, Pakistan





MDLP K13 Ultimate
1 Azerbaijani + Lebanese + North_Ossetian + Pashtun-Afghan @ 1.034991

MDLP K23b
1 Georgian_Jew + Georgian_Jew + Iranian + Uzbek_Afghan @ 2.499159

MDLP World 22
1 Brahui + Lebanese + Tatar_Crim + Turk @ 2.602559

MDLP World
1 Jew_Georgia + Tatar_Crim + Turk + Brahui @ 1.771243

Eurogenes K13
1 Afghan_Pashtun + Azeri + Iranian + Lebanese_Druze @ 2.491446

Eurogenes K15
1 Afghan_Pashtun + Iranian + Kurdish + Lebanese_Druze @ 3.368770

Gedrosia K11
1 Adygei + Lebanese + Makrani + Turkish @ 3.704922

Gedrosia K12
1 Iranian + Iraqi_Arab_Baghdad + Iraqi_Jew + Tajik_Pomiri @ 2.795564

Dodecad V3
1 Assyrian + Brahui + Turks + Turks @ 2.382086

Dodecad K12b
1 Georgia_Jews + Iranians + Tajiks + Uzbekistan_Jews @ 1.793558

Dodecad World 9
1 Assyrian + Iranian + Kurd + Tajiks @ 1.142684

HarappaWorld
1 assyrian + kurd + kurd + tajik @ 1.525492



My map. Although a bit more spread out than yours, note the greatest pin density's in the exact same area as yourself.

Looks like Azeris are basically just plain ol' Median-Oghuz hybrids, eh? :)

http://i.imgur.com/qle6APX.png

My mother's locations. Her map is in yours.

http://oi60.tinypic.com/maugp1.jpg

Tabriz, Urmia, Maragheh, Zanjan, Qazvin, Naqadeh and Takestan are important cities in my mother's map.
Ardabil, Lankaran, Astara, Bilasuvar, Neftchala, Parsabad, Jelilabad, Lahijan, Khalkhal, Zanjan, Qazvin, Rasht and Amol are in mine.

Our results show clearly Azerbaijani populated regions in Iran. Especially my mother's map shows the region predominantly Azerbaijani. Your map is more spread out (to Persian and Kurdish areas) than my and my mom's map because of your half Persian heritage imo.

jesus
10-25-2015, 08:39 PM
http://i.imgur.com/q1hWkc5.png

4 mix oracle Midpoint locations:
Kerman x1
South Khorasan x2
Fars x3
Bushehr x1
Khuzestan x2
Chaharmahal and Bakhtiari x1
Isfahan x1
Lorestan x1

Final midpoint location: Northern Fars province, between Izad Khast and Shurjestan. Pretty accurate since most of my ancestry comes from Fars province and historical Fars.


http://i.imgur.com/bHEtrlc.png


Without GedrosiaDNA calculators, midpoint shifts to central Kohgiluyeh and Boyer-Ahmad

http://i.imgur.com/uL1eRHg.png

Almost all midpoints and the final midpoint placed me in ancient Elam
https://nesu6taylor.files.wordpress.com/2014/07/ane1350.jpg

DMXX
10-25-2015, 10:07 PM
Our results show clearly Azerbaijani populated regions in Iran. Especially my mother's map shows the region predominantly Azerbaijani.

Ah. Just remembered you chose Tabriz for the sample location for the Azeri samples, even though they're the ones from the Republic of Azerbaijan. You and your mother's positions would shift slightly northwards in light of that. I'd assume it'd still place both of you mostly within historical Azerbaijan (south of the Aras) either way. So would I, actually. I haven't seen your 4-Oracle choices, so the extent of the shift's unknown.



Your map is more spread out (to Persian and Kurdish areas) than my and my mom's map because of your half Persian heritage imo.

Indeed, although if my northern Persian ancestry was radically different, I wouldn't be firmly above historical Azerbaijan here. Also looks like there's some distant Azeri on that side, too.

I also think, for this exercise to be more realistic, we'd have to exclude the Jewish results. The differing autosomal profiles between them and their gentile compatriots makes placing them within their countries of origin somewhat problematic here. That's the reason I ended up in Turkmenistan in one of the outputs (Uzbek Jews centred at Uzbekistan).

khanabadoshi
10-26-2015, 01:04 AM
https://scontent.xx.fbcdn.net/hphotos-xfp1/v/t1.0-9/s720x720/11219390_10104100449458268_7612818767045186467_n.j pg?oh=3358e36df4bf5fbf4c49f5d4c059bafe&oe=56D29A4A
https://scontent.xx.fbcdn.net/hphotos-xap1/v/t1.0-9/s720x720/12033039_10104100449508168_7458012086957317490_n.j pg?oh=89c978d0e237c1ca2ce507ef74c8a4a5&oe=56CDCA61

dp
03-09-2016, 12:19 AM
https://scontent.xx.fbcdn.net/hphotos-xfp1/v/t1.0-9/s720x720/11219390_10104100449458268_7612818767045186467_n.j pg?oh=3358e36df4bf5fbf4c49f5d4c059bafe&oe=56D29A4A
https://scontent.xx.fbcdn.net/hphotos-xap1/v/t1.0-9/s720x720/12033039_10104100449508168_7458012086957317490_n.j pg?oh=89c978d0e237c1ca2ce507ef74c8a4a5&oe=56CDCA61

khana,
did you draw a blank :-)
dp

khanabadoshi
03-10-2016, 12:33 PM
khana,
did you draw a blank :-)
dp

I think I deleted the hosted screenshot by accident!

Dante
04-12-2016, 08:36 PM
This is quite interesting. Puts us where the paper trails & family history has pointed.

Though apparently one of my forefathers was born in the Black Sea. :P. Actually how far back does this "map" go? I know, or have very good belief, that a handful of relatives back in the family tree were Jewish [very Jewish surnames] and Jewish DNA "matches" continually pop up for certain gedmatch oracles (as well as 23&me, if low). This is done by using MLPD K23B, Eurogenes K13, V2 K15, and the EUTest which I believe are most accurate for Europeans?

8788 self

8786 sibling

8787 distant cousin, for MLPD K23B "northern european" would be what exactly. I stuck down Norway as that has been mentioned in passing. I know next to nothing about his family & I don't think he knows much more than that.

dp
05-17-2016, 05:49 PM
If I read your hot spots correct your ethnicity is largely Alpine, Med, with maybe some MENA?
dp

This is quite interesting. Puts us where the paper trails & family history has pointed.

Though apparently one of my forefathers was born in the Black Sea. :P. Actually how far back does this "map" go? I know, or have very good belief, that a handful of relatives back in the family tree were Jewish [very Jewish surnames] and Jewish DNA "matches" continually pop up for certain gedmatch oracles (as well as 23&me, if low). This is done by using MLPD K23B, Eurogenes K13, V2 K15, and the EUTest which I believe are most accurate for Europeans?

8788 self

8786 sibling

8787 distant cousin, for MLPD K23B "northern european" would be what exactly. I stuck down Norway as that has been mentioned in passing. I know next to nothing about his family & I don't think he knows much more than that.

Mike_G
05-27-2016, 10:28 PM
My first shot at mapping this out. Apologize for how busy it looks. Yellow star is the % ratio between the two locations. The four pointed stars show the single country guesstimate. I included the top two results from MDLP K23b. Vadim, I think I broke your program. Swedish Saami?

Adoption paperwork says my mother was Dutch/English. Father was Polish. I guess a North Sea parent and a Slav makes a German? 23andme had it as a 60/30 NWE/EE split speculative. It's all somewhat consistent considering my reported background, but I didn't expect to see as much Scandinavian as I did.

9489

Gaku
07-27-2016, 11:44 PM
10665 sibling

10666 sibling

I am more like the second sibling than the other.

Gaku
07-27-2016, 11:45 PM
duplicate post - can someone delete, thanks.

Tomenable
10-18-2016, 03:46 PM
Oracle results would be preferred as they're more specific. I had a hard time deciding where to place the Near Eastern reference site in that above attempt.

You're more than welcome to PM me your Oracle results from several calculators.

Can I PM you some results too? Which calculators are most welcome?

Calas
04-09-2017, 10:30 AM
Surprising results. I never knew I was from the UK.

Myself - black triangle
Dad - blue triangle
Son - green triangle
Mum - red triangle

Seriously I used more than one oracle from the three [Eurogenes, Dodecad, and MDLP] DMXX suggested and tossed on HarappaWorld too to try and force mum & dad further onto the Continent, wasn't happening.
https://i.imgsafe.org/a0a1d27b0d.png


Huh, married a Scandinavian. And know a rather Finnish individual.
https://i.imgsafe.org/a0a0755273.png

SkyKingVik
06-07-2017, 10:37 PM
Greetings!

I'm having a little trouble on knowing which of the mixed mode population sharing I should be using. My most recent locations that I know of from the paper trail is England, Scotland, Ireland and Isle of Man. We suspect Pakistan, Indian, Turkish area on my father's side because of the dark skin and facial structure. I see Kalash and Baloch come up a lot which would mean Pakistan. I apologize, I struggled with understanding the instructions on choosing a mixed mode population due to my Aspergers/Autism.

Can someone please help me and recommend which ones I should choose. Thank you in advance!

MLDP World:

Primary Population (source) Secondary Population (source) Distance
1 85.8% Austrian + 14.2% Latvian_V 0.77
2 67.9% German-North + 32.1% Montenegrin 0.9
3 55.8% Bosnian + 44.2% Norwegian_V 0.92
4 63.7% German_V + 36.3% Croatian 0.95
5 61.1% German-North + 38.9% Serbian 0.97
6 64.1% Polish_V + 35.9% Kosovar 0.98
7 68.9% Croatian + 31.1% Norwegian_V 0.99
8 63.7% German-South + 36.3% Ukrainian-Center 1.03
9 53% German_V + 47% Hungarian 1.04
10 79.4% Swedish + 20.6% Greek_Center 1.04
11 77.4% Swedish + 22.6% Greek_North 1.04
12 69.3% Sorb + 30.7% Kosovar 1.06
13 66.5% German-South + 33.5% Ukrainian-East 1.07
14 69.3% Swedish + 30.7% Gagauz 1.08
15 54.8% Croatian + 45.2% CEU_V 1.09
16 65.4% Swedish + 34.6% Bulgarian 1.09
17 60.6% Polish + 39.4% Kosovar 1.09
18 80.3% Swedish + 19.7% Greek_East 1.09
19 88.7% German_V + 11.3% Latvian_V 1.12
20 86.5% Austrian + 13.5% Ukrainian-West 1.12

MLDP World-22:
Primary Population (source) Secondary Population (source) Distance
1 98.8% German_V (derived) + 1.2% Tlingit (derived) 1.05
2 98.4% German_V (derived) + 1.6% Haida (derived) 1.06
3 99% German_V (derived) + 1% Athabask (derived) 1.06
4 99.2% German_V (derived) + 0.8% North-Amerind (ancestral) 1.07
5 98% German_V (derived) + 2% Tsimsian (derived) 1.09
6 99.2% German_V (derived) + 0.8% Proto-Indo-Iranian (ancestral) 1.1
7 99.1% German_V (derived) + 0.9% Kalash (derived) 1.11
8 99% German_V (derived) + 1% Navajo (derived) 1.11
9 98.9% German_V (derived) + 1.1% Apache (derived) 1.16
10 99.1% German_V (derived) + 0.9% Eskimo (derived) 1.19
11 97.4% German_V (derived) + 2.6% Aleut (derived) 1.2
12 99.3% German_V (derived) + 0.7% Near_East (ancestral) 1.2
13 98.2% German_V (derived) + 1.8% Miwok (derived) 1.2
14 99% German_V (derived) + 1% Saudi (derived) 1.21
15 99% German_V (derived) + 1% Bedouin (derived) 1.22
16 99.1% German_V (derived) + 0.9% Jew_Yemen (derived) 1.22
17 85.9% German-North (derived) + 14.1% Italian-Center (derived) 1.22
18 98.7% German_V (derived) + 1.3% Ste7 (derived) 1.23
19 99.2% German_V (derived) + 0.8% Bra2 (derived) 1.23
20 98.5% German_V (derived) + 1.5% Mexican (derived) 1.23

Eurogenes K13:

Primary Population (source) Secondary Population (source) Distance
1 90% Southwest_English + 10% Tatar 1.18
2 91.8% Southwest_English + 8.2% Chuvash 1.2
3 92.7% Southwest_English + 7.3% Mari 1.48
4 90.9% Southwest_English + 9.1% Erzya 1.53
5 90% Southwest_English + 10% Kargopol_Russian 1.66
6 88% Southwest_English + 12% Ukrainian_Lviv 1.84
7 89.9% Southwest_English + 10.1% Ukrainian_Belgorod 1.9
8 89.8% Southwest_English + 10.2% Southwest_Russian 1.92
9 88.5% Southwest_English + 11.5% Ukrainian 1.97
10 87.5% Southwest_English + 12.5% South_Polish 2
11 90.3% Southwest_English + 9.7% East_Finnish 2.04
12 88.1% Southwest_English + 11.9% Moldavian 2.07
13 89.9% Southwest_English + 10.1% Russian_Smolensk 2.08
14 63.2% Irish + 36.8% South_Dutch 2.09
15 87.5% Southwest_English + 12.5% Croatian 2.1
16 80% Southwest_English + 20% East_German 2.12
17 88.7% Southwest_English + 11.3% Polish 2.12
18 71% Orcadian + 29% South_Dutch 2.14
19 90.2% Southwest_English + 9.8% Estonian_Polish 2.15
20 84.5% Southwest_English + 15.5% Hungarian 2.15

Dodecad V3:
Primary Population (source) Secondary Population (source) Distance
1 92.8% Orkney (1000 Genomes) + 7.2% Sardinian (HGDP) 1.04
2 95.4% CEU (HapMap) + 4.6% French_Basque (HGDP) 1.14
3 93.8% Orcadian (HGDP) + 6.2% Sardinian (HGDP) 1.19
4 92.3% CEU (HapMap) + 7.7% French (HGDP) 1.27
5 90.5% Orkney (1000 Genomes) + 9.5% IBS (1000Genomes) 1.28
6 90.4% Orkney (1000 Genomes) + 9.6% Spanish (Dodecad) 1.3
7 97% CEU (HapMap) + 3% Spaniards (Behar) 1.3
8 97.2% CEU (HapMap) + 2.8% Spanish (Dodecad) 1.31
9 91.8% Orkney (1000 Genomes) + 8.2% North_Italian (HGDP) 1.31
10 89.8% Orkney (1000 Genomes) + 10.2% Portuguese (Dodecad) 1.33
11 91% Orcadian (HGDP) + 9% Portuguese (Dodecad) 1.33
12 90.8% Orkney (1000 Genomes) + 9.2% N_Italian (Dodecad) 1.33
13 97.4% CEU (HapMap) + 2.6% IBS (1000Genomes) 1.33
14 91.6% Orcadian (HGDP) + 8.4% IBS (1000Genomes) 1.34
15 91.6% Orcadian (HGDP) + 8.4% Spanish (Dodecad) 1.35
16 74.4% Argyll (1000 Genomes) + 25.6% French (Dodecad) 1.35
17 88.5% Argyll (1000 Genomes) + 11.5% Spanish (Dodecad) 1.36
18 97.4% CEU (HapMap) + 2.6% Portuguese (Dodecad) 1.36
19 88.1% Argyll (1000 Genomes) + 11.9% Spaniards (Behar) 1.38
20 90.1% Orkney (1000 Genomes) + 9.9% Spaniards (Behar) 1.39

Harappa World:

Primary Population (source) Secondary Population (source) Distance
1 90.2% orcadian (hgdp) + 9.8% ashkenazi (harappa) 1.01
2 93.1% orcadian (hgdp) + 6.9% cypriot (behar) 1.04
3 94.2% orcadian (hgdp) + 5.8% druze (hgdp) 1.08
4 90.8% orcadian (hgdp) + 9.2% ashkenazy-jew (behar) 1.08
5 91.6% orcadian (hgdp) + 8.4% turk-aydin (hodoglugil) 1.09
6 84.9% orcadian (hgdp) + 15.1% romanian-a (behar) 1.15
7 98.2% utahn-white (1000genomes) + 1.8% brahui (hgdp) 1.17
8 98.2% utahn-white (1000genomes) + 1.8% makrani (hgdp) 1.2
9 98.1% utahn-white (1000genomes) + 1.9% balochi (hgdp) 1.21
10 95.6% british (1000genomes) + 4.4% nogai (yunusbayev) 1.22
11 93.8% orcadian (hgdp) + 6.2% lebanese-druze (haber) 1.22
12 93.8% orcadian (hgdp) + 6.2% lebanese-christian (haber) 1.2
13 86.1% orcadian (hgdp) + 13.9% bulgarian (yunusbayev) 1.23
14 95.9% british (1000genomes) + 4.1% turkmen (yunusbayev) 1.27
15 92.9% orcadian (hgdp) + 7.1% turk-kayseri (hodoglugil) 1.28
16 92% orcadian (hgdp) + 8% sephardic-jew (behar) 1.3
17 93.6% orcadian (hgdp) + 6.4% lebanese-muslim (haber) 1.3
18 95.8% british (1000genomes) + 4.2% stalskoe (xing) 1.31
19 98.2% utahn-white (1000genomes) + 1.8% selkup (rasmussen) 1.31
20 96.3% british (1000genomes) + 3.7% kumyk (yunusbayev) 1.32

EthioHelix K10 + French:
Primary Population (source) Secondary Population (source) Distance
1 82.5% french + 17.5% mozabite 1.82
2 72% french + 28% morocco-n 1.86
3 74.7% french + 25.3% moroccans 1.95
4 71.7% french + 28.3% algeria 2.03
5 78% french + 22% sahara-occ 2.17
6 65.6% french + 34.4% libya 5.17
7 77.8% french + 22.2% morocco-s 5.81
8 56% french + 44% egypt 8.42
9 59.1% french + 40.9% egyptans 9.52
10 81.5% french + 18.5% fulani 12.42
11 77.9% french + 22.1% EtA 15.3
12 78.4% french + 21.6% EtT 15.73
13 78.6% french + 21.4% ethiopian-jews 15.95
14 81.7% french + 18.3% EtO 16.96
15 82.9% french + 17.1% hema 17.04
16 79.6% french + 20.4% EtT-P 17.26
17 83.4% french + 16.6% maasai 17.52
18 80.4% french + 19.6% EtA-P 17.52
19 81.1% french + 18.9% EtO-P 17.54
20 84.1% french + 15.9% sandawe 17.82

Eurasia K9 ASI:

Primary Population (source) Secondary Population (source) Distance
1 72.7% Norwegian + 27.3% Albanian 0.73
2 73.9% Scottish + 26.1% Bulgarian 0.89
3 83.3% Scottish + 16.7% Greek 0.96
4 81.9% Scottish + 18.1% Albanian 1
5 74.7% Norwegian + 25.3% Greek 1.09
6 73.6% Norwegian + 26.4% Tuscan 1.1
7 62.8% Norwegian + 37.2% Bulgarian 1.16
8 98.6% Hungarian + 1.4% Karelia_HG 1.2
9 98.6% Hungarian + 1.4% Samara_HG 1.2
10 98.4% Hungarian + 1.6% MA1 1.22
11 82.6% Scottish + 17.4% Tuscan 1.23
12 78.7% Norwegian + 21.3% Sicilian 1.26
13 86.3% Scottish + 13.7% Sicilian 1.27
14 95.6% Hungarian + 4.4% Srubnaya 1.29
15 96.2% Hungarian + 3.8% RISE_baAndrov 1.3
16 94.5% Hungarian + 5.5% RISE_baUne 1.31
17 93.9% Hungarian + 6.1% RISE_baSca 1.37
18 96% Hungarian + 4% RISE_baSin 1.43
19 79.5% Norwegian + 20.5% Maltese 1.45
20 69.2% Czech + 30.8% Bulgarian 1.47

PuntDNAL K12 Modern:
Primary Population (source) Secondary Population (source) Distance
1 84.1% English_South + 15.9% French_South 3.19
2 91.6% English_South + 8.4% Sardinian 3.23
3 80.2% English_South + 19.8% Spanish_Northeast 3.24
4 97.3% Utahn_European + 2.7% Paniya 3.29
5 87% English_South + 13% Basque_Spaniard 3.36
6 86.8% English_South + 13.2% Spanish_Canaries 3.42
7 83.8% English_South + 16.2% Spanish_Southwest 3.43
8 97.5% Utahn_European + 2.5% Hakkipikki 3.48
9 97.1% German_South + 2.9% Paniya 3.49
10 67.7% English_South + 32.3% French 3.52
11 73.2% Irish + 26.8% Spanish_Northeast 3.52
12 88.2% Irish + 11.8% Sardinian 3.57
13 91.6% Utahn_European + 8.4% Basque_Spaniard 3.6
14 64.8% German_North + 35.2% Spanish_Northeast 3.62
15 70.7% Dutch_North + 29.3% French_South 3.64
16 90.4% Utahn_European + 9.6% French_South 3.64
17 84.1% English_South + 15.9% Italian_Bergamo 3.65
18 98.4% Utahn_European + 1.6% Mayan 3.68
19 98.5% Utahn_European + 1.5% Pima 3.69
20 78.8% Irish + 21.2% French_South 3.69