PDA

View Full Version : Revised and Updated "Geographic Population Structure Prediction"



Sein
01-24-2014, 12:00 PM
Has anyone tried the updated site? It's a huge improvement over what they previously offered, and has a lot of interesting stuff to play with. I like how they've made all of the data public. Nice to see the 1000 Genomes samples analyzed on the Geno 2.0 Chip.
http://www.prosapia.org/

They have a "Download" section, but I'm too inept to use anything on it. So I just tried the "Calculator". I recommend it to anyone who has their Geno 2.0 results.

To save everyone some trouble:
Admixture1=Northeast Asian
Admixture2=Mediterranean
Admixture3=Southern African
Admixture4=Southwest Asian
Admixture5=Native American
Admixture6=Oceanian
Admixture7=Southeast Asian
Admixture8=Northern European
Admixture9=Sub-Saharan African

My Geno 2.0 results:
Northeast Asian=4%
Mediterranean=26%
Southern African=0%
Southwest Asian=46%
Native American=2%
Oceanian=0%
Southeast Asian=11%
Northern European=11%
Sub-Saharan African=0%
(My Mediterranean is 25%, but I added one percent, so that the sum is 100%)

When you input numbers, the sum of Admixture 1-9 is supposed to be equal to 1. So I put in:
Admixture1=0.04
Admixture2=0.26
Admixture3=0
Admixture4=0.46
Admixture5=0.02
Admixture6=0
Admixture7=0.11
Admixture8=0.11
Admixture9=0

My location on the World Map, using 1 reference population:

"Based on the data that you have entered our tool computed that your closest population is from the Iran region.
Calculated Latitude of Origin: 32.4192
Calculated Longitude of Origin: 53.6861
The region is marked on the map below as markers."

1264

Using 10 reference populations:

"Based on the data that you have entered our tool computed that your closest population is from the Iran region.
Calculated Latitude of Origin: 34.0001
Calculated Longitude of Origin: 59.742
The region is marked on the map below as markers."

1265

I expected eastern Iran, not too far from home. If we get enough people to try this, we could perhaps attempt something along the lines of the "Spatial Ancestry" thread. Should be fun to compare results between the two.

If anyone understands the "Downloads" section, please do share with us. I just couldn't figure it out (I guess that reflects rather badly on my mental agility). It has a really cool complete work-flow, so you can create something similar using your own Admixture experiments.

DMXX
01-24-2014, 12:51 PM
I ended up in the North Caucasus. Like yourself, added 1% to the Mediterranean to achieve 100%. When using more than one reference population, my point settles in northern Georgia, just east of South Ossetia.


http://imagizer.imageshack.us/v2/800x600q90/812/h6dd.png



Admixture 1 Admixture 2 Admixture 3 Admixture 4 Admixture 5 Admixture 6 Admixture 7 Admixture 8 Admixture 9
0.05 0.48 0 0.32 0 0 0.03 0.12 0
Reference Population: 1

Sein
01-24-2014, 01:10 PM
Cool! A rather interesting result.

I looked at the Iranian results provided in the CSV, they range from 53% "Southwest Asian", to only 32%. Most are around 40%. Does that give any indication as to their geographic origins? Perhaps they are from the south of the country?

DMXX
01-24-2014, 01:21 PM
I have a feeling the Geno 2.0 Iranians are from somewhere central in Iran, perhaps Yazd or thereabouts. Note the Geno 2.0 Iranian average lacks any components modal in Africa, whereas most of the Behar et al. Iranian samples have some degree of African admixture (http://www.harappadna.org/2011/03/iranians/).

If we presume some degree of African ancestry among Iranians points to a southern location, then the Geno 2.0 Iranians fit somewhere between the zone bound by northerners (e.g. yours truly) and the Behar Iranians. Yazd is a guess. Might be from somewhere like Hamadan province (central-west) for all we know.

For the record, my SW Asian score is 32%.

MfA
01-24-2014, 01:22 PM
This is mine using former Eurogenes emulator ''K9b''

http://abload.de/img/desktop_2014_01_24_156udrx.png


Based on the data that you have entered our tool computed your closest population from Iran region.
Calculated Latitude of Origin: 37.9661
Calculated Longitude of Origin: 52.5416
The region is marked on the map below as markers.


0.011 Northeast Asian
0.372 Mediterranean
0.000 Southern African
0.430 Southwest Asian
0.001 Native American
0.000 Oceanian
0.000 Southeast Asian
0.185 Northern European
0.000 Sub-Saharan African

MfA
01-24-2014, 01:29 PM
Can you guys who have already done Geno2 try K9b? would be interesting to see how much it is off.

Sein
01-24-2014, 01:35 PM
I have a feeling the Geno 2.0 Iranians are from somewhere central in Iran, perhaps Yazd or thereabouts. Note the Geno 2.0 Iranian average lacks any components modal in Africa, whereas most of the Behar et al. Iranian samples have some degree of African admixture (http://www.harappadna.org/2011/03/iranians/).

If we presume some degree of African ancestry among Iranians points to a southern location, then the Geno 2.0 Iranians fit somewhere between the zone bound by northerners (e.g. yours truly) and the Behar Iranians. Yazd is a guess. Might be from somewhere like Hamadan province (central-west) for all we know.

For the record, my SW Asian score is 32%.

Sounds very reasonable. It's fascinating that you have the minimum amount of "Southwest Asian" for the Iranians they sampled. I'm sure your results are more reflective of northern Iranians, but in comparison to the Iranians they sampled, you seem more similar to the Ingush, but with an intermediate Northern European percentage (you have more than these Iranians, but less than the Ingush).

soulblighter
01-24-2014, 01:36 PM
My Dad's ( http://www.anthrogenica.com/showthread.php?719-Dad-s-Geno-2-results&p=4492&viewfull=1#post4492 ) results (I added 1% each to the SW asian and SE asian to get to 100%):

One population:
1267


Ten populations:
1268


The download section is for an offline calculator for the same.

Sein
01-24-2014, 01:39 PM
Can you guys who have already done Geno2 try K9b? would be interesting to see how much it is off.

My K9b results:
Southwest_Asian=60.94%
North_European=11.06%
Mediterranean=10.07%
Southeast_Asian=7.20%
Native_American=2.40%
Northeast_Asian=3.70%
Oceanian=3.45%
Sub-Saharan_African=1.18%

Very noisy in comparison to the actual results. Also underestimates "Mediterranean", and overestimates "Southwest Asian". "Northern European" seems accurate though.

DMXX
01-24-2014, 01:42 PM
My K9b results:
Southwest_Asian=60.94%
North_European=11.06%
Mediterranean=10.07%
Southeast_Asian=7.20%
Native_American=2.40%
Northeast_Asian=3.70%
Oceanian=3.45%
Sub-Saharan_African=1.18%

Very noisy in comparison to the actual results. Also underestimates "Mediterranean", and overestimates "Southwest Asian". "Northern European" seems accurate though.

In my case, it's really overestimating my North Euro (usually around 8-12%) as well as the others:



Population
Southwest_Asian 42.33%
Native_American 0.42%
Northeast_Asian 3.53%
Mediterranean 31.23%
North_European 19.15%
Southeast_Asian 1.69%
Oceanian 0.82%
South_African -
Sub-Saharan_African 0.83%


I don't think there's sufficient carry-over here.

Sein
01-24-2014, 01:42 PM
My Dad's ( http://www.anthrogenica.com/showthread.php?719-Dad-s-Geno-2-results&p=4492&viewfull=1#post4492 ) results (I added 1% each to the SW asian and SE asian to get to 100%):

One population:
1267


Ten populations:
1268


The download section is for an offline calculator for the same.

Pretty cool that your father is in Northern India, just like what Dr. McDonald provides for other South Indian Brahmins.

MfA
01-24-2014, 01:52 PM
My K9b results:
Southwest_Asian=60.94%
North_European=11.06%
Mediterranean=10.07%
Southeast_Asian=7.20%
Native_American=2.40%
Northeast_Asian=3.70%
Oceanian=3.45%
Sub-Saharan_African=1.18%

Very noisy in comparison to the actual results. Also underestimates "Mediterranean", and overestimates "Southwest Asian". "Northern European" seems accurate though.


In my case, it's really overestimating my North Euro (usually around 8-12%) as well as the others:



Population
Southwest_Asian 42.33%
Native_American 0.42%
Northeast_Asian 3.53%
Mediterranean 31.23%
North_European 19.15%
Southeast_Asian 1.69%
Oceanian 0.82%
South_African -
Sub-Saharan_African 0.83%


I don't think there's sufficient carry-over here.

Thanks, as it expected it looks quite off.. I ran DMXX's results for the sake of comparison.


Based on the data that you have entered our tool computed your closest population from Tajikistan region.
Calculated Latitude of Origin: 37.96
Calculated Longitude of Origin: 63.9307
The region is marked on the map below as markers.

http://glacombio01.comp.glam.ac.uk/projects/ancestry/record.php?jobID=24012014-014747-5713



http://abload.de/img/desktop_2014_01_24_15lgiv2.png

soulblighter
01-24-2014, 01:52 PM
Pretty cool that your father is in Northern India, just like what Dr. McDonald provides for other South Indian Brahmins.


That area near madhya pradesh has some significance, even in our religious history:

"Narmadayai namah pratah Narmadayai namo nisi

Namostu Narmade tubhyam pahi mam visa-sarpatah"

Is one of the verses recited daily during religious rituals by my ethnic group (http://en.wikipedia.org/wiki/Vadama). I always wondered why the river Narmada had a special significance. May be it is pure coincidence . As much as I think religion is B.S, there may be history woven into all the myth and rituals.

emmental
01-24-2014, 02:42 PM
Can you guys who have already done Geno2 try K9b? would be interesting to see how much it is off.

I used K9b and it put me around Stockholm, Sweden. My ancestry (to about 350ybp) is approx. 60% Swiss, 30% Dutch/North German and 10% South German.

Seems off to me.

Sein
01-24-2014, 02:44 PM
That area near madhya pradesh has some significance, even in our religious history:

"Narmadayai namah pratah Narmadayai namo nisi

Namostu Narmade tubhyam pahi mam visa-sarpatah"

Is one of the verses recited daily during religious rituals by my ethnic group (http://en.wikipedia.org/wiki/Vadama). I always wondered why the river Narmada had a special significance. May be it is pure coincidence . As much as I think religion is B.S, there may be history woven into all the myth and rituals.

That's quite an intriguing connection.

Here are some results for a North Indian Brahmin (roots in Uttar Pradesh, and Uttarakhand):

"Based on the data that you have entered our tool computed your closest population from Gujarati region.
Calculated Latitude of Origin: 26.7362
Calculated Longitude of Origin: 74.2889
The region is marked on the map below as markers.

You may wish to take a note of result reference 24012014-023429-3704 to see the analysis in future."

1269

Results for an Afghan Pashtun friend of mine:

"Based on the data that you have entered our tool computed your closest population from Iran region.
Calculated Latitude of Origin: 35.425
Calculated Longitude of Origin: 59.0188
The region is marked on the map below as markers.

You may wish to take a note of result reference 24012014-121337-3506 to see the analysis in future."

1270

I like how they end up in virtually the same exact spot as myself. They actually have some Afghan Hazara ancestry, about 12.5%. Yet despite this, still rather similar to myself.

Sein
01-24-2014, 02:46 PM
I used K9b and it put me around Stockholm, Sweden. My ancestry (to about 350ybp) is approx. 60% Swiss, 30% Dutch/North German and 10% South German.

Seems off to me.

It seems the K9b model doesn't really approximate Geno 2.0 very well.

Dr_McNinja
01-24-2014, 02:47 PM
I got the same as that North Indian Brahmin, somewhere in Rajasthan (says Gujarat):


Based on the data that you have entered our tool computed your closest population from Gujarati region.
Calculated Latitude of Origin: 27.2455
Calculated Longitude of Origin: 74.5057

That was with 10 reference populations.

With 9 or fewer (and greater than 1) reference populations, it's a little closer to the border:


Calculated Latitude of Origin: 27.4715
Calculated Longitude of Origin: 72.9214My Geno 2 results:



50% Southwest Asian
19% Southeast Asian
16% Mediterranean
11% Northern European
2% Native American
2% Northeast Asian

Dr_McNinja
01-24-2014, 02:50 PM
It sounds like they were doing something similar to what I did in the spreadsheets (averaging latitude/longitude of admixture components), but with the few components used, it's not really distinguishing people within an ethnic cluster too well.

EDIT: Here's the map I got from porting over all the Harappa Ancestry Project participants: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk

Sein
01-24-2014, 02:53 PM
I got the same as that North Indian Brahmin, somewhere in Rajasthan (says Gujarat)...

Your results are strikingly similar to the North Indian Brahmin:
Southwest Asian=50%
Southeast Asian=19%
Mediterranean=17%
Northern European=12%
Native American=2%
This individual has roots in western Uttarakhand, and northwestern Uttar Pradesh.

Sein
01-24-2014, 02:55 PM
It sounds like they were doing something similar to what I did in the spreadsheets (averaging latitude/longitude of admixture components), but with the few components used, it's not really distinguishing people within an ethnic cluster too well.

EDIT: Here's the map I got from porting over all the Harappa Ancestry Project participants: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk

A very good point. But I think they also used the actual reference population coordinates. They put them in a CSV.

Dr_McNinja
01-24-2014, 03:04 PM
Your results are strikingly similar to the North Indian Brahmin:
Southwest Asian=50%
Southeast Asian=19%
Mediterranean=17%
Northern European=12%
Native American=2%
This individual has roots in western Uttarakhand, and northwestern Uttar Pradesh.Here's a Punjabi Jatt Sikh for comparison (paul gill I believe):


Southwest Asian 50%
Mediterranean 19%
Southeast Asian 15%
Northern European 13%
Northeast Asian 2%It's like the Brahmin is between me and him, except for the Northeast Asian which is curiously absent in the Brahmin but present in both of us.

EDIT: Comparing myself (HRP0349) and this Brahmin (HRP0380 I presume?), my results seem a lot noisier.

HRP0380:


S-Indian: 35.94
Baloch: 35.50
Caucasian: 10.77
NE-Euro: 10.00
SE-Asian: 0
Siberian: 0.90
NE-Asian: 0
Papuan: 0.73
American: 0.64
Beringian: 0.74
Mediterranean: 4.42
SW-Asian: 0.02
San: 0
E-African: 0
Pygmy: 0
W-African: 0Me:
S-Indian: 34.97
Baloch: 33.49
Caucasian: 10.69
NE-Euro: 8.34
SE-Asian: 0.73
Siberian: 2.25
NE-Asian: 0
Papuan: 1.61
American: 1.67
Beringian: 2.17
Mediterranean: 1.24
SW-Asian: 2.16
San: 0.30
E-African: 0
Pygmy: 0.38
W-African: 0Calculator effect in... effect?

Sein
01-24-2014, 03:16 PM
That's rather interesting. Considering how endogamous Jatts are, I think the Northeast Asian component is a Scythian contribution (I guess it can't be Turkic, so Scythian makes more sense).

Paul Gill ends up very close to the Indo-Pak border:

"Based on the data that you have entered our tool computed your closest population from Gujarati region.
Calculated Latitude of Origin: 28.1814
Calculated Longitude of Origin: 72.775
The region is marked on the map below as markers.

You may wish to take a note of result reference 24012014-122647-9602 to see the analysis in future."

1271

Sein
01-24-2014, 03:29 PM
Calculator effect in... effect?

Lol, I like that. To be honest though, HRP0380 seems quite different from the Metspalu UP-Brahmin and UP-Kshatriya. More Punjabi-like. Is it because they are from northwestern UP, with some substantial western Uttarkhandi ancestry?

But I absolutely agree. HarappaWorld is long due for an overhaul. The results seem to get noisier and noisier, with more participants.

Dr_McNinja
01-24-2014, 04:34 PM
Lol, I like that. To be honest though, HRP0380 seems quite different from the Metspalu UP-Brahmin and UP-Kshatriya. More Punjabi-like. Is it because they are from northwestern UP, with some substantial western Uttarkhandi ancestry?

But I absolutely agree. HarappaWorld is long due for an overhaul. The results seem to get noisier and noisier, with more participants.Are there any easy guides/tutorials to using the admixture software? I'd love to at least play around with it on my own.

Sein
01-24-2014, 04:46 PM
Are there any easy guides/tutorials to using the admixture software? I'd love to at least play around with it on my own.

Here's Razib Khan's guide.
http://blogs.discovermagazine.com/gnxp/2011/03/analyzing-ancestry-with-admixture-step-by-step/

vettor
01-24-2014, 04:56 PM
Interesting

The proto/beta 1st time this was done ....about a year ago it gave me central france ................you are a dutch .......

The next proper which was same time as above , it gave me ..........western Hungaria .............you are a lithuanian ......

this current is
Southwest_Asian 11.96%
Native_American 0.57%
Northeast_Asian -
Mediterranean 29.66%
North_European 55.93%
Southeast_Asian 1.41%
Oceanian 0.24%
South_African -
Sub-Saharan_African 0.23%

Based on the data that you have entered our tool computed your closest population from Dane region.
Calculated Latitude of Origin: 54.8616
Calculated Longitude of Origin: 8.92146

Placed on the border of Denmark and Germany

3 tries...all could be correct .........who can tell

A dane..........maybe a cimbrian, that came from denmark , fought the Romans in the alps, lost and hid in the alps for the next 2000 years

ZephyrousMandaru
01-24-2014, 04:57 PM
I don't know how to compute my ADMIXTURE Proportions using this calculator.

Eurogenes K9b

Population
Southwest_Asian 35.80%
Native_American -
Northeast_Asian -
Mediterranean 44.77%
North_European 19.15%
Southeast_Asian -
Oceanian 0.25%
South_African -
Sub-Saharan_African -

My Northern European is highly overestimated, it's typically in the single and sometimes I have none at all.

Sein
01-24-2014, 05:00 PM
I don't know how to compute my ADMIXTURE Proportions using this calculator.

Eurogenes K9b

Population
Southwest_Asian 35.80%
Native_American -
Northeast_Asian -
Mediterranean 44.77%
North_European 19.15%
Southeast_Asian -
Oceanian 0.25%
South_African -
Sub-Saharan_African -

My Northern European is highly overestimated, it's typically in the single and sometimes I have none at all.

If you want, I could do it for you. Just paste your actual Geno 2.0 results here (the K9b will give results that are rather off).

vettor
01-24-2014, 05:10 PM
I don't know how to compute my ADMIXTURE Proportions using this calculator.

Eurogenes K9b

Population
Southwest_Asian 35.80%
Native_American -
Northeast_Asian -
Mediterranean 44.77%
North_European 19.15%
Southeast_Asian -
Oceanian 0.25%
South_African -
Sub-Saharan_African -

My Northern European is highly overestimated, it's typically in the single and sometimes I have none at all.

i got this for you

Based on the data that you have entered our tool computed your closest population from Ingush region.
Calculated Latitude of Origin: 42.8283
Calculated Longitude of Origin: 44.0333

border of russia and georgia...terek valley

ZephyrousMandaru
01-24-2014, 05:13 PM
i got this for you

Based on the data that you have entered our tool computed your closest population from Ingush region.
Calculated Latitude of Origin: 42.8283
Calculated Longitude of Origin: 44.0333

border of russia and georgia...terek valley

Thanks, how did you do it?

Sein
01-24-2014, 05:14 PM
Having found your actual results:
"Based on the data that you have entered our tool computed your closest population from Georgians region.
Calculated Latitude of Origin: 38.74
Calculated Longitude of Origin: 42.175
The region is marked on the map below as markers."

1278

Interesting to compare with the K9b results. I think using the actual Geno 2.0 results is much more accurate.

parasar
01-24-2014, 05:14 PM
Lol, I like that. To be honest though, HRP0380 seems quite different from the Metspalu UP-Brahmin and UP-Kshatriya. More Punjabi-like. Is it because they are from northwestern UP, with some substantial western Uttarkhandi ancestry?

But I absolutely agree. HarappaWorld is long due for an overhaul. The results seem to get noisier and noisier, with more participants.

Brahmans from Uttarkhand tend to show some Tibeto-Burman influence (there was one Uttaranchal Brahmin in the Metspalu dataset pg 11 http://download.cell.com/AJHG/mmcs/journals/0002-9297/PIIS0002929711004885.mmc1.pdf ).

HRP380 is a Tyagi (agriculturist), which is akin to my caste. Tyagis are settled in western UP, Uttarkhand, Haryana, Himachal, and eastern Punjab. Their counterparts in the Western Punjab and Afghanistan used to be the Mohyals.

Tyagis (Taga) are from the plains and according to their accounts are supposed to have originated in Bengal/Gaur. The Jats of the region around Delhi over time proved to be better farmers than Tyagis, and the Jats became the main agricultural caste in the past few hundred years.
Eg. of landholding patterns in the Akbar period: http://books.google.com/books?id=-nMIAAAAQAAJ&pg=PA3
The Taga in the tables are the Tyagi.

The legends of the Punjab have indications of this movement from Bengal.
The Legend of Raja Gopi Chand
'The hero's country is, however, said to be Gaur BangAla or Bengal"
http://books.google.com/books?id=ulcIAAAAQAAJ&pg=PA1
"Kahte Shahr Ujjain Rao nit karte bhog bilasa.
Gaur Bangala, des jinhon ka tyag dia biswasa."

ZephyrousMandaru
01-24-2014, 05:20 PM
Can someone enter my actual Geno 2.0 ADMIXTURE values into this calculator? They're 59% Mediterranean, 29% Southwest Asian and 9% Northern European. Thanks.

Sein
01-24-2014, 05:22 PM
Can someone enter my actual Geno 2.0 ADMIXTURE values into this calculator? They're 59% Mediterranean, 29% Southwest Asian and 9% Northern European. Thanks.

The map at #31 is the result using your actual values for Geno 2.0.

Humanist
01-24-2014, 05:24 PM
Having found your actual results:
"Based on the data that you have entered our tool computed your closest population from Georgians region.
Calculated Latitude of Origin: 38.74
Calculated Longitude of Origin: 42.175
The region is marked on the map below as markers."

1278

Interesting to compare with the K9b results. I think using the actual Geno 2.0 results is much more accurate.

So, Zeph is right next to Lake Van, it appears? This goes back to something I posted in the other thread. There is a problem when populations like Assyrians and Iraqi Mandaeans plot so far to the north. This is, of course, a function of the limited reference populations (among other problems). When Armenians and Georgians are appearing as our most similar populations, these results are inevitable.

Sein
01-24-2014, 05:29 PM
So, Zeph is right next to Lake Van, it appears? This goes back to something I posted in the other thread. There is a problem when populations like Assyrians and Iraqi Mandaeans plot so far to the north. This is, of course, a function of the limited reference populations. When Armenians and Georgians are appearing as our most similar populations, these results are inevitable.

An excellent point, they really need far more robust sampling from across West Asia. But I guess it's accurate, considering the reference populations. Georgians are the least East Asian of all the Caucasus populations they have, and Zeph has zero East Asian, so it finds the best fit?

ZephyrousMandaru
01-24-2014, 05:29 PM
So, Zeph is right next to Lake Van, it appears? This goes back to something I posted in the other thread. There is a problem when populations like Assyrians and Iraqi Mandaeans plot so far to the north. This is, of course, a function of the limited reference populations. When Armenians and Georgians are appearing as our most similar populations, these results are inevitable.

I was just thinking the same. My Eurogenes K9b results are even more northern shifted, I think it might be due to a combination of an inadequate reference set and my overestimated Northern European admixture.

vettor
01-24-2014, 05:45 PM
If geno people test with geno markers and plot and then test with K9b markers and plot, we could actually measure the trend of the program, which to means seems to overestimate a more southern ( head to africa) bias.

Sein
01-24-2014, 05:57 PM
If geno people test with geno markers and plot and then test with K9b markers and plot, we could actually measure the trend of the program, which to means seems to overestimate a more southern ( head to africa) bias.

Let me give this a try.

"Based on the data that you have entered our tool computed your closest population from Gujarati region.
Calculated Latitude of Origin: 27.1876
Calculated Longitude of Origin: 73.2903
The region is marked on the map below as markers."

Went from Iran to Rajasthan, because of the Southwest Asian component.

everest59
01-24-2014, 06:43 PM
Here's Razib Khan's guide.
http://blogs.discovermagazine.com/gnxp/2011/03/analyzing-ancestry-with-admixture-step-by-step/

That's a good link. He has two different references. The one called phylocore is a bigger file with 130k snp's.
Some programming is needed if you want to do a supervised run.
Zack's website is also very helpful. He has a bunch of codes there.
I mainly use Perl, which is pretty easy.
As for admixtools, it's even easier and a lot faster.
Main thing is, learn to convert to various formats.

Dr_McNinja
01-24-2014, 07:44 PM
With K9B:

Population
Southwest_Asian 64.29%
Native_American 2.53%
Northeast_Asian 1.12%
Mediterranean 3.35%
North_European 8.38%
Southeast_Asian 12.64%
Oceanian 4.80%
South_African -
Sub-Saharan_African 2.90%(Added 1% to Med, N-Euro, NE-Asian, and SE-Asian... )

Based on the data that you have entered our tool computed your closest population from Gujarati region.
Calculated Latitude of Origin: 24.1138
Calculated Longitude of Origin: 75.2943That's a bit to the southeast from before.

Sein
01-25-2014, 12:11 AM
Found Biloo's results. Biloo is 3/4 Algerian, and 1/4 Ottoman Turk from Western Anatolia:

"Based on the data that you have entered our tool computed your closest population from Tunisia region.
Calculated Latitude of Origin: 36.1914
Calculated Longitude of Origin: 11.0012
The region is marked on the map below as markers.

You may wish to take a note of result reference 25012014-120850-7620 to see the analysis in future."

1280

Rather accurate. East and north of Algeria though, but I think the 25% Turkish ancestry explains that.

Sein
01-25-2014, 01:38 AM
"I have a feeling the Geno 2.0 Iranians are from somewhere central in Iran, perhaps Yazd or thereabouts. Note the Geno 2.0 Iranian average lacks any components modal in Africa, whereas most of the Behar et al. Iranian samples have some degree of African admixture."

You're absolutely spot on, I just checked the actual coordinates, it seems these samples are from Yazd Province. All I can say, you're awesome.

toast
01-25-2014, 02:56 AM
I got the same as that North Indian Brahmin, somewhere in Rajasthan (says Gujarat):

That was with 10 reference populations.

With 9 or fewer (and greater than 1) reference populations, it's a little closer to the border:

My Geno 2 results:

Im confused if you guys are plotting rajasthan then why it says gujarati. Are there no rajastani samples on there at all?

DMXX
01-25-2014, 02:57 AM
So, Zeph is right next to Lake Van, it appears? This goes back to something I posted in the other thread. There is a problem when populations like Assyrians and Iraqi Mandaeans plot so far to the north. This is, of course, a function of the limited reference populations (among other problems). When Armenians and Georgians are appearing as our most similar populations, these results are inevitable.

With this in mind, I decided to experiment a little with select results if we took ZephyrousMandaru's result and compared it with where he should *probably* lie (placed in triangle between Mosul, Duhok and Arbil).

By comparing the two, and then replicating the changes across other results, a shift in placement may be observed. The end result is a latitude reduction by 2.29 and a longitude addition by 1.19. In short, users are positioned slightly further south and east of their original placements.

Here are some of the findings. I didn't want to complicate the multiple results posted by other users. Anyone can make a similar position fudging using "the Zephyrous factor". :)



Latt (T) Long (T) Latt (A) Long (A) Location (A), description
Zeph 38.74 42.18 36.45 43.37 N Iraq (reference)
Sein 34 59.74 31.71 60.93 Afghan-Iranian border
DMXX 43.42 45.72 41.13 46.91 N Azerbaijan
N Ind Brahmin 26.74 74.29 24.45 75.48 Central-NW India
Afgh Pashtun 35.43 59.02 33.14 60.21 E Iran (near Afghan border)


(T) = Theoretical, (A) = "Actual"


"I have a feeling the Geno 2.0 Iranians are from somewhere central in Iran, perhaps Yazd or thereabouts. Note the Geno 2.0 Iranian average lacks any components modal in Africa, whereas most of the Behar et al. Iranian samples have some degree of African admixture."

You're absolutely spot on, I just checked the actual coordinates, it seems these samples are from Yazd Province. All I can say, you're awesome.

My pleasure!

Sein
01-25-2014, 03:38 AM
...With this in mind, I decided to experiment a little with select results if we took ZephyrousMandaru's result and compared it with where he should *probably* lie (placed in triangle between Mosul, Duhok and Arbil).

By comparing the two, and then replicating the changes across other results, a shift in placement may be observed. The end result is a latitude reduction by 2.29 and a longitude addition by 1.19. In short, users are positioned slightly further south and east of their original placements...

Excellent! Quite an improvement, and I like how me and the Afghan Pashtun are virtually identical, but with them having a slight shift towards Central Asia, consistent with their 12.5% Hazara ancestry. Rather accurate.

I decided to co-opt our friend Imadaddin's Geno 2.0 results.

"Based on the data that you have entered our tool computed your closest population from Abkhazia region.
Calculated Latitude of Origin: 38.3259
Calculated Longitude of Origin: 45.5588
The region is marked on the map below as markers.

You may wish to take a note of result reference 25012014-032452-1497 to see the analysis in future."

1281

Somewhat to your south. Your position on the map seems to be consistently in the Caucasus.

Humanist
01-25-2014, 05:57 AM
With this in mind, I decided to experiment a little with select results if we took ZephyrousMandaru's result and compared it with where he should *probably* lie (placed in triangle between Mosul, Duhok and Arbil).

A learned man you are, my friend. That is a very good approximation of the modern Assyrian Heartland (c. 1st millennium CE). The ancient Assyrian Heartland was Nineveh (~Mosul), Arbil, and Assur (c. 1st millennium BCE).

vettor
01-25-2014, 06:33 AM
[QUOTE=DMXX;28454]With this in mind, I decided to experiment a little with select results if we took ZephyrousMandaru's result and compared it with where he should *probably* lie (placed in triangle between Mosul, Duhok and Arbil).

/QUOTE]

How do you know where he should probably be?............from what he says or?

This GPS system............what is its suppose oldest date range for each member...500 years......1000 years

I am confused in the sense I do not know how people can actually say they are from this spot on the globe no matter what.

Maybe I need some understanding of this GPS system and what people expect from it

regards

Humanist
01-25-2014, 07:05 AM
With this in mind, I decided to experiment a little with select results if we took ZephyrousMandaru's result and compared it with where he should *probably* lie (placed in triangle between Mosul, Duhok and Arbil).

How do you know where he should probably be?............from what he says or?

DMXX is obviously taking the collective record into consideration.

1. ZephyrousMandaru is from a village in modern northern Iraq, in what is today Kurdistan.

2. ZephyrousMandaru's vernacular is the "definitely endangered" language of Sureth, a form of Eastern Aramaic native to Mesopotamia, and with possible influences from the Assyrian variety of Akkadian (see the work of Cambridge professor Geoffrey Khan, and the Russian linguist Sergei Loesov).

3. ZephyrousMandaru is a member of the Church of the East. Two of the major centers of the Church of the East were Seleucia-Ctesiphon (near Babylon) and Arbil, the former being the seat of the Patriarch in the 1st millennium CE. The area around modern Duhok was also a significant part of our Church's history.

4. E Assyrians are most similar to Iraqi Mandaeans (former speakers of Mandaic, a "critically endangered" Eastern Aramaic language related to Babylonian Aramaic), and Eastern Mizrahim Jews.

Stephen1986
01-25-2014, 12:39 PM
I'm 45% Northern European, 37% Mediterranean and 17% Southwest Asian. My ancestors are mostly from North West England, western Wales, Gloucestershire, Suffolk, Hertfordshire, County Meath, County Dublin and County Mayo. My first reference population is British, my second is German.

With one reference population and adding 1% to NE -

Based on the data that you have entered our tool computed your closest population from UK region.
Calculated Latitude of Origin: 51.9991
Calculated Longitude of Origin: -3.16593

This is in south west Wales.

One reference population and adding 1% to Med -

Based on the data that you have entered our tool computed your closest population from German region.
Calculated Latitude of Origin: 52.5177
Calculated Longitude of Origin: 13.3586

This is in Berlin.

One reference population and adding 1% to SWA -

Based on the data that you have entered our tool computed your closest population from UK region.
Calculated Latitude of Origin: 52.0118
Calculated Longitude of Origin: -3.1746

This is again in Wales, but slightly more north western.

With ten reference populations and adding 1% to NE -

Based on the data that you have entered our tool computed your closest population from UK region.
Calculated Latitude of Origin: 52.0685
Calculated Longitude of Origin: -2.06535

In the southern Midlands region of England, halfway between Bristol and Birmingham.

With ten reference populations and adding 1% to Med -

Based on the data that you have entered our tool computed your closest population from German region.
Calculated Latitude of Origin: 52.4635
Calculated Longitude of Origin: 11.1342

Further west this time, almost to Wolfsburg.

With ten reference populations and adding 1% to SWA -

Based on the data that you have entered our tool computed your closest population from UK region.
Calculated Latitude of Origin: 52.0768
Calculated Longitude of Origin: -1.60336

Further east this time.

DMXX
01-25-2014, 02:55 PM
I decided to co-opt our friend Imadaddin's Geno 2.0 results.
...
Somewhat to your south. Your position on the map seems to be consistently in the Caucasus.

Curious positioning. I would've expected our results to be the other way round considering most of my ancestors are from further south of his. I do have some ancestry from the North Caucasus on my paternal side, but it's not recent according to my grandfather's research (6-7 generations ago). Perhaps my "excessively" northern position on various charts (including McDonald's BGA) indicates it was more recent than that?

Or, this is (for whatever reason) an expected result for a north Iranian with my form of results. If NK19191 has done Geno 2.0 or K9b, that'd be a great indication.

NK19191
01-25-2014, 03:47 PM
@DMXX

I haven't done Geno 2.0, but below is my K9b results.

Eurogenes K9b Admixture Proportions

Population
Southwest_Asian 43.92%
Native_American -
Northeast_Asian 2.96%
Mediterranean 31.31%
North_European 18.59%
Southeast_Asian 2.05%
Oceanian 0.41%
South_African 0.28%
Sub-Saharan_African 0.48%



Admixture1=Northeast Asian= 0.0296
Admixture2=Mediterranean=0.3131+1=0.3231
Admixture3=Southern African= 0.0028
Admixture4=Southwest Asian=0.4392
Admixture5=Native American-0.0
Admixture6=Oceanian=0.0041
Admixture7=Southeast Asian=0.00205
Admixture8=Northern European=0.1859
Admixture9=Sub-Saharan African=0.0048

using 10 Reference Population:

GPS Calculation
Based on the data that you have entered our tool computed your closest population from Tajikistan region.
Calculated Latitude of Origin: 37.3043
Calculated Longitude of Origin: 60.9655
The region is marked on the map below as markers.

You may wish to take a note of result reference 25012014-040427-2094 to see the analysis in future.


http://i1309.photobucket.com/albums/s632/nk191919/Mapofmylocation_zpsae455cde.png

I wonder if I did this right?????

DMXX
01-25-2014, 04:41 PM
I wonder if I did this right?????

I only opined for your K9b results so I could compare my K9b location with yours. From that, we would get an idea of where your Geno 2.0 plot would be by transposing the difference (as did using "the Zephyrous factor" in the previous page).

My K9b spot (entered exact numbers like you, made up .82% deficit by adding to Mediterranean) is latitude 38.3535, longitude 68.487. I sit just west of Dushanbe in Tajikistan.

I think you needed to add a bit more than just 1 to your Med score to come to "1", as you arrive at 0.9915 even with that addition. I've added an extra 0.0085 to your Med score to balance that out a bit. Your spot ends up being around 38.3431, 68.4907. Nearly identical to mine.

So, if you had Geno 2.0, you'd end up approximately where I do, which is somewhere between the North Caucasus and Georgia. The recurring problem here is that you also have some known ancestry from that part of West Eurasia. There isn't any clear indication whether this position is normal for north Iranians, or whether our Caucasian heritage is shifting us north, or whether we're just unusually northern.

newtoboard
01-25-2014, 04:46 PM
How reliable is K9b? It seems to be overestimating how Med south Asians are and how Northern European West Asians are.

vettor
01-25-2014, 04:52 PM
I only opined for your K9b results so I could compare my K9b location with yours. From that, we would get an idea of where your Geno 2.0 plot would be by transposing the difference (as did using "the Zephyrous factor" in the previous page).

My K9b spot (entered exact numbers like you, made up .82% deficit by adding to Mediterranean) is latitude 38.3535, longitude 68.487. I sit just west of Dushanbe in Tajikistan.

I think you needed to add a bit more than just 1 to your Med score to come to "1", as you arrive at 0.9915 even with that addition. I've added an extra 0.0085 to your Med score to balance that out a bit. Your spot ends up being around 38.3431, 68.4907. Nearly identical to mine.

So, if you had Geno 2.0, you'd end up approximately where I do, which is somewhere between the North Caucasus and Georgia. Problem here again, is, you also have some known ancestry from that part of West Eurasia. There isn't any clear indication whether this position is normal for north Iranians, or whether our Caucasian heritage is shifting us north, or whether we're just unusually northern.

thanks

so, I have no geno2.0 and use k9b , but I have 2 accounts ( one ftdna and one 23andme). The difference is 80km further east globally for ftdna results

my ftdna results
http://imageshack.com/scaled/medium/34/7xqo.png (http://imageshack.com/photo/my-images/34/7xqo.png/) Uploaded with ImageShack.us (http://imageshack.com)

both my maternal and paternal lineages go back to around 1700 in North-East Italy ( not via word, but all via birth, death and marriage registras )
My question is why is there a big gap from Denmark to NE Italy for me ..................does this GPS sytem calculate ancient, medieval or renaissance period for each individual.

Also, have you found a big difference between geno2.0 and k9b positions globally ..................some say 1000km error is K9b, an error heading in a NW direction


kindest regards

DMXX
01-25-2014, 04:53 PM
How reliable is K9b? It seems to be overestimating how Med south Asians are and how Northern European West Asians are.

For the purposes of this thread, not very. It very broadly correlates with the Geno 2.0 results only.

It seems to form clusters in a different manner to other calculators (even those from Eurogenes). Note that Zephyrous, NK19191 and myself all have ~19% North Euro on K9b, yet in other calculators, Zephyrous has minimal/no North European, whereas NK19191 and me are usually in the 8-12% range.


thanks
so, I have no geno2.0 and use k9b , but I have 2 accounts ( one ftdna and one 23andme). The difference is 80km further east globally for ftdna results


Using my results as a reference for a moment, to convert a K9b position to Geno 2.0, it'll be latitude -5, longitude +22.5. I am doubtful this is a gold standard by any means considering the issues described above; K9b handles components in a different way, so one-to-one correlation shouldn't be expected.

ZephyrousMandaru
01-25-2014, 05:53 PM
I'm honored to be part of this science experiment. LOL But what I don't understand is, if Eastern Assyrians originated in Southeastern Iraq (Babylon). Then why is it that we plot further north today? What exactly causing is this recent northern shift?

Joe B
01-25-2014, 06:39 PM
thanks

so, I have no geno2.0 and use k9b , but I have 2 accounts ( one ftdna and one 23andme). The difference is 80km further east globally for ftdna results

my ftdna results
http://imageshack.com/scaled/medium/34/7xqo.png (http://imageshack.com/photo/my-images/34/7xqo.png/) Uploaded with ImageShack.us (http://imageshack.com)

both my maternal and paternal lineages go back to around 1700 in North-East Italy ( not via word, but all via birth, death and marriage registras )
My question is why is there a big gap from Denmark to NE Italy for me ..................does this GPS sytem calculate ancient, medieval or renaissance period for each individual.

Also, have you found a big difference between geno2.0 and k9b positions globally ..................some say 1000km error is K9b, an error heading in a NW direction


kindest regards
Using K9b with 23andme data put me just offshore of Stockholm which is about 1000 km north of where the Eurogenes Oracles generally plot. It just may not work as well for some European populations.

Hanna
01-25-2014, 06:48 PM
This is what I got using Eurogenes K9b

http://imageshack.com/a/img577/2039/d7rl.jpg

Southwest_Asian 35.13%
Native_American -
Northeast_Asian 4.86%
Mediterranean 35.78%
North_European 24.23%
Southeast_Asian -
Oceanian -
South_African -
Sub-Saharan_African


This is what my aunt got. Is it normal for her to be near Kuwait!?

http://imageshack.com/a/img842/2548/z53k.jpg

Southwest_Asian 37.33%
Native_American 0.17%
Northeast_Asian 0.55%
Mediterranean 40.03%
North_European 21.62%
Southeast_Asian -
Oceanian 0.29%
South_African -
Sub-Saharan_African

DMXX
01-25-2014, 06:58 PM
^ You might want to double-check you've put the right ancestral components in, Hanna. I ended up in Kuwait as well initially before realising I entered my North Euro score into Component 9 (SSA).

Hanna
01-25-2014, 07:08 PM
^ You might want to double-check you've put the right ancestral components in, Hanna. I ended up in Kuwait as well initially before realising I entered my North Euro score into Component 9 (SSA).

Oh I did the same mistake. Here is my aunt's new location:

http://imageshack.com/a/img15/7084/1suw.jpg

MitchellSince1893
01-25-2014, 07:32 PM
I agree with others that using K9b isn't a good option. I normally come out in either Denmark, Northern Germany, Netherlands, or the North Sea area where these 3 countries are nearest.

Using this calculator I came out 600 miles Northwest of these areas, on the West Coast of Finland.

Calculated Latitude of Origin: 61.3191
Calculated Longitude of Origin: 21.5137

This is obviously way off based on my ancestry.

Anglecynn
01-25-2014, 07:37 PM
I agree with others that using K9b isn't a good option. I normally come out in either Denmark, Northern Germany, Netherlands, or the North Sea area where these 3 countries are nearest.

Using this calculator I came out 600 miles Northwest of these areas, on the West Coast of Finland.

Calculated Latitude of Origin: 61.3191
Calculated Longitude of Origin: 21.5137

This is obviously way off based on my ancestry.

What order do you enter the components in? I came out in Mexico with this, obviously wrong. I entered them in this order
1 = Southwest_Asian
2 = Native_American
3 = Northeast_Asian
4 = Mediterranean
5 = North_European
6 = Southeast_Asian
7 = Oceanian
8 = South_African
9 = Sub-Saharan_African

ZephyrousMandaru
01-25-2014, 07:41 PM
Oh I did the same mistake. Here is my aunt's new location:

http://imageshack.com/a/img15/7084/1suw.jpg

Hanna, it's best to take these with a fraction of a grain of salt. The lack of reference populations is what's pushing you very far north.

Imadaddin Nasimi
01-25-2014, 07:46 PM
My Geno 2.0 location = Iranian Azerbaijan

http://i.imgur.com/pd4eAuI.png

Well done! Successful prediction. I am Azerbaijani (Iranian Azerbaijani and Republic Azerbaijani mixed)

Geno 2.0 admixture here:
http://i.imgur.com/eLoRJ9J.png


And my Eurogenes K9b location (used FTDNA autosomal Raw Data) = Turkmen

http://i.imgur.com/4AZaREJ.png

K9b results

Southwest_Asian 35%
Native_American 1%
Northeast_Asian 8%
Mediterranean 34%
North_European 19%
Southeast_Asian 3%

NK19191
01-25-2014, 08:01 PM
^ It puts you somewhere around Northwest of Tabriz near city of Marand maybe.

For those who don't know Marand is a city between Tabriz and Khoy in East Azerbaijan Province.

vettor
01-25-2014, 08:02 PM
My Geno 2.0 location = Iranian Azerbaijan

http://i.imgur.com/pd4eAuI.png

Well done! Successful prediction. I am Azerbaijani (Iranian Azerbaijani and Republic Azerbaijani mixed)

Geno 2.0 admixture here:
http://i.imgur.com/eLoRJ9J.png


And my Eurogenes K9b location (used FTDNA autosomal Raw Data) = Turkmen

http://i.imgur.com/4AZaREJ.png

K9b results

Southwest_Asian 35%
Native_American 1%
Northeast_Asian 8%
Mediterranean 34%
North_European 19%
Southeast_Asian 3%

and which is closer to you?

should eurogenes k9b be taken as legit or not...........i mean there is over 10% difference in some of your numbers

vettor
01-25-2014, 08:07 PM
What order do you enter the components in? I came out in Mexico with this, obviously wrong. I entered them in this order
1 = Southwest_Asian
2 = Native_American
3 = Northeast_Asian
4 = Mediterranean
5 = North_European
6 = Southeast_Asian
7 = Oceanian
8 = South_African
9 = Sub-Saharan_African

from post #1
Admixture1=Northeast Asian
Admixture2=Mediterranean
Admixture3=Southern African
Admixture4=Southwest Asian
Admixture5=Native American
Admixture6=Oceanian
Admixture7=Southeast Asian
Admixture8=Northern European
Admixture9=Sub-Saharan African


you have it wrong

also, I was told you , for the last box, use the number whic is the numbe rof boxes filled, so if you have some with zero , do not count them............makes very little difference in my opinion

Humanist
01-25-2014, 08:12 PM
I'm honored to be part of this science experiment. LOL But what I don't understand is, if Eastern Assyrians originated in Southeastern Iraq (Babylon). Then why is it that we plot further north today? What's exactly causing this recent northern shift?

The Eurogenes SPA results, from what I recall, were intended to give one an idea of very deep (combined) ancestral origins. Without a more "recent" northern shift, the purported Assyrian-Akkadian elements in our vernacular would not make much sense*. Also, are you not familiar with the Good Book? ;p

Genesis

He [Nimrod] was a mighty hunter before the LORD; that is why it is said, Like Nimrod, a mighty hunter before the LORD. The first centers of his kingdom were Babylon, Uruk, Akkad and Kalneh, in Shinar. From that land he went to Assyria, where he built Nineveh, Rehoboth Ir, Calah and Resen, which is between Nineveh and Calah--which is the great city.

*
Neo-Babylonian, Late Babylonian, and Standard Babylonian are attested long after Neo-Assyrian writings ceased.

Imadaddin Nasimi
01-25-2014, 08:18 PM
and which is closer to you?

should eurogenes k9b be taken as legit or not...........i mean there is over 10% difference in some of your numbers

both.

My locations from Dodecad Globe13 Palisto's Distance Calculator

first,
http://i.imgur.com/fX5RFAv.png
latitude: 40.05 longitude:51.34

I am a fish in the Caspian sea :)

second,
http://i.imgur.com/3EuVfOU.png
latitude: 40.68 longitude:53.47


And my location from DMXX's Location Calculator with my Gedmatch oracle results. From this topic: http://www.anthrogenica.com/showthread.php?1778-Mapping-Your-Location-(GEDmatch-calculators)

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

All of locations are similar. I think I am typical Caspian Man.;) with western Caspian and eastern Caspian genes.

Anglecynn
01-25-2014, 08:22 PM
from post #1
Admixture1=Northeast Asian
Admixture2=Mediterranean
Admixture3=Southern African
Admixture4=Southwest Asian
Admixture5=Native American
Admixture6=Oceanian
Admixture7=Southeast Asian
Admixture8=Northern European
Admixture9=Sub-Saharan African


you have it wrong

also, I was told you , for the last box, use the number whic is the numbe rof boxes filled, so if you have some with zero , do not count them............makes very little difference in my opinion

Thanks for the information. :)

It doesn't seem to work with K9b, last time i got Lithuanian, this time Finnish:

Based on the data that you have entered our tool computed your closest population from Finnish region.
Calculated Latitude of Origin: 60.9959
Calculated Longitude of Origin: 21.1611
The region is marked on the map below as markers.

Hanna
01-25-2014, 08:35 PM
Hanna, it's best to take these with a fraction of a grain of salt. The lack of reference populations is what's pushing you very far north.

Well it's not too off also. I would have expected my aunt to be located somewhere around Georgia or Armenia.

vettor
01-25-2014, 08:41 PM
my sister

on the swiss -german border near zurich

Based on the data that you have entered our tool computed your closest population from German region.
Calculated Latitude of Origin: 47.4975
Calculated Longitude of Origin: 8.7888

MitchellSince1893
01-25-2014, 08:42 PM
from post #1
Admixture1=Northeast Asian
Admixture2=Mediterranean
Admixture3=Southern African
Admixture4=Southwest Asian
Admixture5=Native American
Admixture6=Oceanian
Admixture7=Southeast Asian
Admixture8=Northern European
Admixture9=Sub-Saharan African


you have it wrong

also, I was told you , for the last box, use the number whic is the numbe rof boxes filled, so if you have some with zero , do not count them............makes very little difference in my opinion

Yes, the order above was the one I used that placed me in Western Finland.

Humanist
01-25-2014, 08:43 PM
I'm honored to be part of this science experiment. LOL But what I don't understand is, if Eastern Assyrians originated in Southeastern Iraq (Babylon). Then why is it that we plot further north today? What exactly causing is this recent northern shift?

Also, I should add, the more southern Eurogenes SPA results may be, at least in part, due to this:


I asked a question a few months ago, on another forum, regarding the possible genetic and linguistic impact on the Assyrian Heartland, in the 8th and 7th centuries BCE, and beyond, by this fact:

Origin points for ~90% of deportations to the Assyrian Heartland.

Font size = % reported in Roads and Mass Deportations in the Neo-Assyrian Empire (David Danzig 2011)

http://i1096.photobucket.com/albums/g326/dok101/assyrian_deported_heartland__.jpg


And, the contents of a related post:

Again, from the Danzig paper, including some possible ideas regarding languages spoken in the regions in question (added by me).

Sources for all Neo-Assyrian deportations, to all destinations, in descending order of frequency:

56.7% Southern Zagros / Elam and Babylonia (Elamite, Akkadian, Iranian*, Indo-Iranian??)
18.4% Middle and South Levant (Canaanite related languages (i.e. ancient Hebrew), NW Semitic (i.e. Old Aramaic))
8.0% Anatolia (W/C Anatolian Indo-European (e.g. Luwian, Hittite), Hurrian?**)
5.0% Northern Zagros and Foothills (Hurro-Urartian?, Akkadian, Iranian*, Indo-Iranian (e.g. Median?))
4.5% North Levant / Upper Euphrates Elbow (NW Semitic (i.e. Old Aramaic), W/C Anatolian Indo-European (e.g. Luwian, Hittite), Hurrian?)
3.5% Lakes Van and Urmia (Hurro-Urartian related languages, E Anatolian IE (proto-Armenian or Armenian?))
2.0% Euphrates and Tigris Sources (Hurro-Urartian related languages?, E Anatolian IE (proto-Armenian or Armenian?), W/C Anatolian Indo-European?? (e.g. Luwian, Hittite))
2.0% Habur Area / Jazira (NW Semitic (i.e. Old Aramaic), Akkadian, Hurrian?)

* Not Indo-Iranian. Or, at least not from what I have been able to understand from the record.
** This may be a bit late for an actual "Hurrian." Not late for a "Hurro-Urartian" related language, of course (i.e. Urartian).

ZephyrousMandaru
01-25-2014, 08:58 PM
The Eurogenes SPA results, from what I recall, were intended to give one an idea of very deep (combined) ancestral origins. Without a more "recent" northern shift, the purported Assyrian-Akkadian elements in our vernacular would not make much sense*. Also, are you not familiar with the Good Book? ;p

Genesis


*
Neo-Babylonian, Late Babylonian, and Standard Babylonian are attested long after Neo-Assyrian writings ceased.

Yeah I know, but what I was asking was, and maybe I should have been more specific, is if whether or not this northern shift is due to ancestry and if so, what ancestry in particular?

MitchellSince1893
01-25-2014, 09:28 PM
double post...see post below

MitchellSince1893
01-25-2014, 09:30 PM
For what it's worth, entering my Eurogenes Hunter_Gatherer Admixture values gave a location that was very close to where I'm placed by Eurogenes K13 and K15.

As I had 0% for Pymy and Bantu, I didn't have to worry about how to split them between Southern African and Sub-Saharan African. I did have a small amount of East African Pastorialist and I put this in the Southern African field. When I placed it in the Sub-Saharan African field I came out in Northern Jutland. It seemed to be a better match if I left it in the Southern African field.

I entered these Hunter-Gather values for the calculator:
Admixture1=Northeast Asian: East Asian Farmer
Admixture2=Mediterranean: Mediterranean Farmer
Admixture3=Southern African: Try, East African Pastoralist, Pymy or Bantu or some combo
Admixture4=Southwest Asian: Added Anatolian Farmer and Middle Eastern Herder
Admixture5=Native American: South American Hunter Gatherer
Admixture6=Oceanian: Oceanian Hunter Gatherer
Admixture7=Southeast Asian: South Asian Hunter Gatherer (I know this isn't an ideal fit but I had to put it somewhere)
Admixture8=Northern European: Added Baltic Hunter Gatherer and North Eurasian Hunter Gatherer
Admixture9=Sub-Saharan African: Try, East African Pastoralist, Pymy or Bantu or some combo

As I said previously this is very close to my Eurogene K13 (Denmark near Copenhagen) and K15 results (N Germany North East of Hamburg near Baltic coast).

The Eurogenes Hunter_Gatherer values kinda split the difference between my K13 and K15 locations.
1284

I'd be interested to know how the Eurogenes Hunter_Gatherer values come out for you all.

vettor
01-25-2014, 09:50 PM
For what it's worth, entering my Eurogenes Hunter_Gatherer Admixture values gave a location that was very close to where I'm placed by Eurogenes K13 and K15.

As I had 0% for Pymy and Bantu, I didn't have to worry about how to split them between Southern African and Sub-Saharan African. I did have a small amount of East African Pastorialist and I put this in the Southern African field. When I placed it in the Sub-Saharan African field I came out in Northern Jutland. It seemed to be a better match if I left it in the Southern African field.

I entered these Hunter-Gather values for the calculator:
Admixture1=Northeast Asian: East Asian Farmer
Admixture2=Mediterranean: Mediterranean Farmer
Admixture3=Southern African: Try, East African Pastoralist, Pymy or Bantu or some combo
Admixture4=Southwest Asian: Added Anatolian Farmer and Middle Eastern Herder
Admixture5=Native American: South American Hunter Gatherer
Admixture6=Oceanian: Oceanian Hunter Gatherer
Admixture7=Southeast Asian: South Asian Hunter Gatherer (I know this isn't an ideal fit but I had to put it somewhere)
Admixture8=Northern European: Baltic Hunter Gatherer
Admixture9=Sub-Saharan African: Try, East African Pastoralist, Pymy or Bantu or some combo

As I said previously this is very close to my Eurogene K13 (Denmark near Copenhagen) and K15 results (N Germany North East of Hamburg near Baltic coast).

The Eurogenes Hunter_Gatherer values kinda split the difference between my K13 and K15 locations.
1284

I'd be interest to know how the Eurogenes Hunter_Gatherer values come out for you all.

your system

Based on the data that you have entered our tool computed your closest population from Bulgaria region.
Calculated Latitude of Origin: 44.2205
Calculated Longitude of Origin: 21.1894

on the border.........serbia, romania and bulgaria

always a possibility as northern Italians sit near bulgarians in most admixture plots

Sein
01-25-2014, 11:17 PM
Curious positioning. I would've expected our results to be the other way round considering most of my ancestors are from further south of his. I do have some ancestry from the North Caucasus on my paternal side, but it's not recent according to my grandfather's research (6-7 generations ago). Perhaps my "excessively" northern position on various charts (including McDonald's BGA) indicates it was more recent than that?

Or, this is (for whatever reason) an expected result for a north Iranian with my form of results. If NK19191 has done Geno 2.0 or K9b, that'd be a great indication.

This is rather interesting. Imadaddin's results seem to be quite accurate, so I'm assuming this boils down to sampling. If they had sampled northern Iranians, perhaps your results would be far more congruent with your actual geographic origins? This holds for me and my Afghan Pashtun friend. With no samples from the northwestern Pakistani border or Afghanistan, we consistently cluster in central-eastern Iran. And for our northwest Indian/Indo-Aryan Pakistani friends, the lack of northwest South Asian populations means clustering in Rajasthan or Gujarat.

Although, the latter isn't technically inaccurate, since Pashtuns speak an Iranic language, and live on the Iranian plateau, while Punjabi people are Indo-Aryans who live on the northwestern edge of the sub-continent. Fits nicely with history and language.

DMXX
01-25-2014, 11:26 PM
^ I think the underlying reason why there's a mismatch between my/NK19191's spot and our geographical placement compared to Imadaddin is because of the component scores we have rather than sampling.

The "best fit" populations are definitely limited, I do agree there. However, this population structure position program relies entirely on K=9 component values to make its' placements. Comparing my Geno 2.0 results with Imadaddin, I'm slightly more SW Asian (3%) and N Euro (4%) but less Mediterranean (3%) and NE Asian (5%). I would've imagined those differences would've resulted in a negligible net movement, given all four peak in different directions. I can only surmise, based on this, a different "weight" is attributed to each component.

Sein
01-25-2014, 11:56 PM
^ I think the underlying reason why there's a mismatch between my/NK19191's spot and our geographical placement compared to Imadaddin is because of the component scores we have rather than sampling.

The "best fit" populations are definitely limited, I do agree there. However, this population structure position program relies entirely on K=9 component values to make its' placements. Comparing my Geno 2.0 results with Imadaddin, I'm slightly more SW Asian (3%) and N Euro (4%) but less Mediterranean (3%) and NE Asian (5%). I would've imagined those differences would've resulted in a negligible net movement, given all four peak in different directions. I can only surmise, based on this, a different "weight" is attributed to each component.

That probably explains it.

I wonder if anyone with technical expertise in Admixture (everest59 :)) could perhaps try something out using their instructions (copied and pasted):

"A complete Workflow for using GPS on your data
-----------------------------------------------

This workflow explains how we have prepared the input files for the GPS algorithm.
If you have a dataset for which you wish to run GPS, follow this worklow.
Please note, this workflow was tested only on human subjetcs.

--------------------------------
First stage (carried only once):
--------------------------------
Start from a PLINK .ped file (with M SNPs and N samples) that contains samples from as diverse ethnicities as possible with known geographical coordiantes
and at least 12 individuals from each ethnicity and a comparable .map file. For example:

DiverseSamples.ped:
African African 0 0 0 0 a c g t c a
African African 0 0 0 0 a c g t c a
Asian Asian 0 0 0 0 a c g t a a
Oceanian Oceanian 0 0 0 0 c c t t c a
European European 0 0 0 0 c c t t c a
SouthAmerican SouthAmerican 0 0 0 0 a c t t a a

DiverseSamples.map
1 rs324 0 90900
1 rs32344 0 91900
1 rs32454 0 100900


Next, run ADMIXTURE (Alexander, Novembre, and Lange 2009) in an unsupervised mode with Ks starting at 3 and increasing in 1.
ADMIXTURE would produce two output files containing the q a p matrices. You can plot the results of your analyses (for the q file) using tools like
CLUMPP (http://www.stanford.edu/group/rosenberglab/clumpp.html) and observe the results. Find the K in which ethnicities would be homogeneous within themselves
but heterogeneous with other ethnicities. Heterogeneity within samples may suggest either over segmentation or a genetically diverse ethnicity and the decision
should be made based on the known demographic history of each population.

Find the appropriate Kx, which satisfies the above condition and look at the p matrice which has M SNPs and Kx columns. Each column corresponds to a different
hypothestical ancestral population. The p matrice would have the allele frequencies of that hypothestical ancestral populations at each SNP.
Next, Generate the genotypes of individual samples in each hypothestical ancestral population based on the allele frequencies.
For example, for allele frequency of 0.5, we can generate CCCCCCCCCCGGGGGGGGGG and using random sampling, five individuals would have the genotypes:
CC CG CG CC CG CC CG GG GG GG. We suggest generating at least 15 simulated individuals for each hypothestical ancestral population.

Next, plot your original dataset with the hypothetical populations on a PCA plot and examine whether the hypothestical populations cluster where expected.
Hypothestical populations that closely cluster each other may be removed. Your final hypothestical populations should be considered "putative ancestral populations".

From your diverse dataset, select the populations that exhibit the highest homogeneity within themselves for Kx.
The admixture proportions of these populations would make up the GEN matrix.
The geographical coordinates of these populations would make up the GEO matrix.

----------------------------------------------
Second stage (carried for every test samples):
----------------------------------------------
Provided a test sample in a file in PLINK format, compile a .ped file (PLINK format) which includes the test sample and the putative ancestral populations.
For example:

.ped file
---------
test test 0 0 0 0 a a a c c c
Ancestral1 Ancestral1 0 0 0 0 a a a c c g
Ancestral1 Ancestral1 0 0 0 0 a a a c c g
Ancestral1 Ancestral1 0 0 0 0 a a a c c g
Ancestral2 Ancestral1 0 0 0 0 a t c c c g
Ancestral2 Ancestral2 0 0 0 0 a t c c g g
Ancestral2 Ancestral2 0 0 0 0 a t c c g g
Ancestral3 Ancestral3 0 0 0 0 a a a a c c
Ancestral3 Ancestral3 0 0 0 0 a t a a c c
Ancestral3 Ancestral3 0 0 0 0 t t a a c c

Run ADMIXTURE in a supervised mode and mark the putative ancestral populations in the .pop file. For example:

.pop file
---------
-- space for the test sample --
Ancestral1
Ancestral1
Ancestral1
Ancestral2
Ancestral2
Ancestral2
Ancestral3
Ancestral3
Ancestral3

The program would infer the admixture components of the test sample, which would be relative to the putative ancestral populations.

Next, use the GPS algorithm to find the geographical location of the test sample.
The GPS algorithm accepts as input the GEN and GEO matrices along with the admixture components of the test sample and outputs the geographical coordinates
of the test sample along with the most closely related population."

Edit: Reading this again (I seriously need to improve my reading comprehension skills), it seems the actual reference population coordinates provided in the GEO CSV do play some role in the output.

Dr_McNinja
01-26-2014, 01:20 AM
^ Wish there was some way to use that on Windows. Or would it work with Ubuntu booted off a USB?

Sein
01-26-2014, 01:28 AM
^ Wish there was some way to use that on Windows. Or would it work with Ubuntu booted off a USB?

I think the latter is possible, but only everest59 would know from experience.

Stephen1986
01-26-2014, 02:52 AM
I'd be interest to know how the Eurogenes Hunter_Gatherer values come out for you all.

My HG results -

Baltic HG 56.8%
Med Farmer 33.6%
Anat Farmer 7.62%
EAf Pastoralist 0.73%
SAs HG 0.5%
North Eura HG 0.28%
Oceanian HG 0.2%
SoAm HG 0.16%
Pygmy HG 0.11%

Myself with one reference population -

Based on the data that you have entered our tool computed your closest population from Dane region.
Calculated Latitude of Origin: 56.015
Calculated Longitude of Origin: 9.98778

This comes out at around Skanderborg.

Myself with ten reference populations -
Based on the data that you have entered our tool computed your closest population from Dane region.

Calculated Latitude of Origin: 54.4622
Calculated Longitude of Origin: 7.34353

This is in the North Sea, but south of the Danish-German border.

My brother's HG results -

Baltic HG 54.98%
Med Farmer 32.86%
Anat Farmer 8.42%
EAf Pastoralist 0.47%
SAs HG 2.46%
SoAm HG 0.59%
ME Herder 0.15%

He seems to get the same results as I do.

everest59
01-26-2014, 03:10 AM
It doesn't work on Windows. I don't know about booting off USB. You should try it.
I basically followed Razib's instructions, which is very good. So, basically, download the Wubi application. That's basically like an addon to Windows. It is what I did.
I am however in the process of creating a bigger dataset. Once that's complete (been a little busy lately), I'll definitely try to use the software above. I like the fact that the instructions above teaches how to create zombies (if I understood it correctly).
I think what Razib forgot to do with his dataset is to convert builds. For example, some datasets use Build 37 (or hg 19), while others use Build 36 or hg 18. I found a nice software that does it for me.
IMO, anybody can run these softwares. There are a lot of open-source scripts out there. You don't even need to know programming that much.

Anglecynn
01-26-2014, 03:15 AM
For what it's worth, entering my Eurogenes Hunter_Gatherer Admixture values gave a location that was very close to where I'm placed by Eurogenes K13 and K15.

As I had 0% for Pymy and Bantu, I didn't have to worry about how to split them between Southern African and Sub-Saharan African. I did have a small amount of East African Pastorialist and I put this in the Southern African field. When I placed it in the Sub-Saharan African field I came out in Northern Jutland. It seemed to be a better match if I left it in the Southern African field.

I entered these Hunter-Gather values for the calculator:
Admixture1=Northeast Asian: East Asian Farmer
Admixture2=Mediterranean: Mediterranean Farmer
Admixture3=Southern African: Try, East African Pastoralist, Pymy or Bantu or some combo
Admixture4=Southwest Asian: Added Anatolian Farmer and Middle Eastern Herder
Admixture5=Native American: South American Hunter Gatherer
Admixture6=Oceanian: Oceanian Hunter Gatherer
Admixture7=Southeast Asian: South Asian Hunter Gatherer (I know this isn't an ideal fit but I had to put it somewhere)
Admixture8=Northern European: Added Baltic Hunter Gatherer and North Eurasian Hunter Gatherer
Admixture9=Sub-Saharan African: Try, East African Pastoralist, Pymy or Bantu or some combo

As I said previously this is very close to my Eurogene K13 (Denmark near Copenhagen) and K15 results (N Germany North East of Hamburg near Baltic coast).

The Eurogenes Hunter_Gatherer values kinda split the difference between my K13 and K15 locations.
1284

I'd be interest to know how the Eurogenes Hunter_Gatherer values come out for you all.

Here's what i got:
Based on the data that you have entered our tool computed your closest population from Dane region.
Calculated Latitude of Origin: 54.5426
Calculated Longitude of Origin: 6.57116

Closer than before.

My Grandfather:
Based on the data that you have entered our tool computed your closest population from Dane region.
Calculated Latitude of Origin: 54.4665
Calculated Longitude of Origin: 7.25961
The region is marked on the map below as markers.

Grandmother:

Based on the data that you have entered our tool computed your closest population from Dane region.
Calculated Latitude of Origin: 54.0477
Calculated Longitude of Origin: 5.30793
The region is marked on the map below as markers.

It gets them right relative to each other, but not closest overall. My grandmother is generally a close match to SW English and my grandfather to Dutch/Danish.

LUKE33
05-02-2014, 03:52 PM
http://www.youtube.com/watch?v=iMAtIZ6Y1P0#t=37

http://prosapiagenetics.com/ Company web-site

http://prosapiagenetics.com/community/viewforum.php?f=2 forum


hmmm looks like open source stuff dolled up IMO

Decide for yourselves.

Clinton P
05-02-2014, 04:01 PM
Read Dienekes' Anthropology Blogs, here (http://dienekes.blogspot.co.uk/2014/05/the-geographic-position-structure-gps.html) and here (http://dienekes.blogspot.co.uk/2014/04/nature-communications-genographic.html).

Don't waste your money, it's complete nonsense.

Clinton P

LUKE33
05-02-2014, 04:24 PM
Dienekes said...

@Eran Elhaik

Your paper uses at least two ideas that were first published by me: converting unsupervised ADMIXTURE ones into supervised ones via "zombies"; testing a population's similarity to a reference panel by calculating Euclidean distance over the space of admixture coefficients and finding the closest matching population.

Prior work should be cited when it presents a method that is used in the current work. For example, you cite ADMIXTURE which is a component in your current work. You should have done the same for the ideas used in your paper that were previously published by myself.

You can argue independent invention, but this is hard to believe given the timeline, the known readership of my blog, and the fact that my ideas have spread beyond it and have been used by other genome bloggers and third party tools have been developed around them.

In any case, even if you came up with these concepts independently, it is still proper form to cite prior work that is relevant (which this clearly is). And, if you really "were informed of your work only at the reviewer stage and acknowledgment was not allowed", then you can always write a letter to the editor acknowledging the prior publication of part of your method.


http://dienekes.blogspot.co.uk/2014/04/nature-communications-genographic.html

Clinton P
05-02-2014, 04:30 PM
There are a number of cases where their analysis places them in the sea, as in this example here (http://prosapiagenetics.com/results.php?order=613RF45lb1362). :crazy:

Clinton P

LUKE33
05-02-2014, 05:05 PM
Found this also about Eran Elhaik

http://forward.com/articles/175912/jews-a-race-genetic-theory-comes-under-fierce-atta/?p=all

AJL
05-02-2014, 05:15 PM
Elhaik's "theories" have already been debunked elsewhere in this forum.

LUKE33
05-02-2014, 05:18 PM
Elhaik's "theories" have already been debunked elsewhere in this forum.

Just mentioning for the benefit of the newbie...better they have some background info on him.

vettor
05-02-2014, 07:51 PM
There are a number of cases where their analysis places them in the sea, as in this example here (http://prosapiagenetics.com/results.php?order=613RF45lb1362). :crazy:

Clinton P

There are small islands there like the dutch colonized one

virgin islands - was danish
Anguilla - was english
saint Martin - was english
sint Maarten - was dutch

vettor
05-02-2014, 07:56 PM
Found this also about Eran Elhaik

http://forward.com/articles/175912/jews-a-race-genetic-theory-comes-under-fierce-atta/?p=all

This came about due to the fact that judism is only 6000 years old, so all markers before cannot be jewish. People need to be more exacting in naming a subclade which can be jewish, that is, less than 6000 years of age

vettor
05-02-2014, 08:08 PM
It seems like the same company as a year ago with a different name.

there beta test had me in central France and 2 months later there true test placed me in Hungaria

ADW_1981
05-02-2014, 08:14 PM
There are small islands there like the dutch colonized one

saint Martin - was english


This was actually French, not English.

George Chandler
05-02-2014, 10:35 PM
I see this is being offered through Prosapia Genetics for $119 but that gives you the "basic" GPS test yet the example showing results is for "super" GPS testing..which doesn't seem to be an option on the products page.

Does anyone know anything about this?

George

George Chandler
05-02-2014, 10:43 PM
The test offers basic GPS testing but the example for results show super GPS testing which isn't identified in the product line? Is this sort of like showing how 111 STR's reveal much and then offer a test with the same name but only 12 markers?

George

thetick
05-03-2014, 02:14 AM
Seems at best a very poor project and at worst a scam. Checkout all the unhappy customers:

http://prosapiagenetics.com/community/viewtopic.php?f=2&t=7

And follow the link to the "Expert Corner". A scientist who worked on the implementation was supposed to be answering question today... just crickets chirping. Save you money for FTDNA and 23andme tests.

DebbieK
05-03-2014, 04:04 PM
Please see my blog post about Prosapia Genetics:

http://cruwys.blogspot.co.uk/2014/05/driving-in-wrong-direction-with-dodgy.html

AJL
05-03-2014, 04:26 PM
Anyone who is thinking of buying this service -- you can do it yourself free, and better, using this thread:

http://www.anthrogenica.com/showthread.php?1778-Mapping-Your-Location-(GEDmatch-calculators)

Agamemnon
05-03-2014, 04:37 PM
This came about due to the fact that judism is only 6000 years old, so all markers before cannot be jewish. People need to be more exacting in naming a subclade which can be jewish, that is, less than 6000 years of age

Judaism certainly isn't 6,000 years old, the Jewish people's roots stretch back to c. 1200 BCE (so 3,200 years ago), Judaism in its current form is much younger.

Humanist
05-03-2014, 05:37 PM
Elhaik's "theories" have already been debunked elsewhere in this forum.

Agreed. Who can forget the Armenians as a proxy for Khazars thing. :\

Though, the one thing I do like about this fella is his willingness to utilize social media to engage a wider audience, rather than simply fellow academics via the print publication route.


http://youtu.be/Aap-s1kle4Q

warwick
05-03-2014, 10:05 PM
Essentially, the founder of that organization has certain fundamental credibility problems, which can be connected to his endorsement of the views of Arthur Koestler in a scientific paper.

AJL
05-03-2014, 10:37 PM
I have merged three related threads into this one.

Sein
05-04-2014, 12:36 AM
I think the trouble here lies in the very conceptual fabric of this sort of analysis. The assumptions just don't work. They just position genotypes in relation to reference populations, sort of like PCA, but in the space of ADMIXTURE proportions. The only difference, they use the actual geographic coordinates of the populations, and position you in relation to those coordinates. So, the results make sense in the "big" genetic picture, but rarely if ever will they actually pinpoint your "ancestral homeland" (as if there can only be one "ancestral homeland"). Isolated Sardinian villagers don't count (anyway, that's in the data-set they optimized this analysis on, so it's not even a fair comparison).

For example, this tool puts me in northeastern Iran. Not necessarily wrong. In fact, this result makes good sense, as I am a sort of "northeastern" Iranian. Most tools that try to pinpoint a single geographic location for my genotype put me in eastern Iran, or western Afghanistan. This is fairly accurate in terms of my genetic ancestry. Also, two Afghan Pashtuns get the same exact location as myself, despite being from Central Afghanistan. Again, fairly accurate, because we are all from a single ethnic group. But, this is extremely inaccurate when it comes to my actual ancestral "homeland", the Afghanistan-Pakistan borderlands, or for my Afghan Pashtun friends, who are from Central Afghanistan. Mashad is very far away from Ghazni or Peshawar. There simply isn't a clean, unambiguous relationship between one's actual geographic origins, and one's biogeographic origins, which is why a tool like this just can't do what it claims. Although, it does confirm one's genetic affinity to various global populations. It tells me that I'm most similar to their Iranian population, and that my Punjabi friends are most similar to their Gujarati population, and if that's something you want to learn, it's good at that. But, we already have GEDMatch, which is free, and has more calculators and reference populations (the Iranian result for myself, and the Gujarati result for Punjabis, is really due to the lack of reference populations from the eastern Iranian plateau and northwestern South Asia). Also, the free version of this "GPS" is still available. If you've taken the Geno 2.0 test, just input your results into the "demo" calculator. No need to spend money on this.

Humanist
05-04-2014, 02:43 AM
Please see my blog post about Prosapia Genetics:

http://cruwys.blogspot.co.uk/2014/05/driving-in-wrong-direction-with-dodgy.html

DebbieK wrote (http://cruwys.blogspot.co.uk/2014/05/driving-in-wrong-direction-with-dodgy.html):


[T]here are many free services that you can use to get an alternative reading of your data and a prediction of your "ethnicity", all of which will give much better results than the commercial offerings from Prosapia.

Based on what I have read about it, that certainly appears to be the case.

This bit also caught my attention:


Despite the fact that the Prosapia Genetics website appears to have been launched on the same day that the paper was published neither Eran Elhaik nor Tatiana Tatarinova, who are also the two lead authors, have declared any "competing financial interests".

Stellaritic
05-07-2014, 12:02 AM
The groundbreaking Geographic Population Structure (GPS) tool, created by Dr Eran Elhaik from the University of Sheffield's Department of Animal and Plant Sciences and Dr Tatiana Tatarinova from the University of Southern California, works similarly to a satellite navigation system as it helps you to find your way home, but not the one you currently live in – but rather your actual ancestor's home from 1,000 years ago.

Previously, scientists have only been able to locate where your DNA was formed to within 700km, which in Europe could be two countries away; however this pioneering technique has been 98 per cent successful in locating worldwide populations to their right geographic regions, and down to their village and island of origin


Dr Elhaik's co-author, Dr Tatiana Tatarinova, developed a website making GPS accessible to the public.

"To help people find their roots, I developed a website that allows anyone who has had their DNA genotyped to upload their results and use GPS to find their ancestral home," said Dr Tatarinova, who is also an Associate Professor of Research Paediatrics at the Keck School of Medicine of the University of Southern California.

"We were surprised by the simplicity and precision of this method. People in a given geographical area are more likely to have similar genetics. When they also have genetic traits typically found in other, distant regions, the geographical origin of those traits is generally the closest location where those traits can be found."


I just stumbled upon this , I did some Googling but the website was nowhere to be found.
Does anybody know where I can upload my genotype rawdata ?

https://www.youtube.com/watch?v=Aap-s1kle4Q#t=27

http://www.sheffield.ac.uk/news/nr/dna-sat-nav-gps-tool-find-your-ancestors-home-1.370846




------------------update-----------

http://chcb.saban-chla.usc.edu/gps/index.php

GailT
05-07-2014, 12:57 AM
You should read Debbie blog post (link (http://cruwys.blogspot.co.uk/2014/05/driving-in-wrong-direction-with-dodgy.html)) before paying money for this "GPS" analysis. Several people who paid through Paypal were able to get their money refunded.

redifflal
11-20-2014, 01:51 PM
Crap, I used this and paid the 43 bucks and got garbage results. I should have heeded the warnings in this thread. Where is the free demo calculator?

everest59
11-21-2014, 01:33 AM
All that these programs do is use a programming language like R and create a linear model. For South Asians, it is going to be very difficult to calculate coordinates due to the caste system. The only way to do it is to ask the user's caste or ethnicity and then calculate the coordinates.
Also, accuracy depends on the training data used.
I could write few lines of codes to do this. It's a huge waste of money.

jesus
01-26-2015, 07:59 AM
Description:

Geographic Population Structure (GPS) is a biogeographic tool that analyzes your autosomal DNA data and predicts your most recent geographical origin. GPS has a time resolution of up to 1,000 years back. Unlike other genomics-based ancestry technologies, GPS does not use Mitochondrial or Y chromosomal haplogroups because they provide a limited geographic resolution of the past 15,000-20,000 years.

In plain English, what does GPS do?

GPS finds a point of origin for an individual, such as a village, town or country, where one's DNA was formed by combination of several gene pools from populations coming together and creating a genetic line that would eventually lead to the individual being analyzed. Historic examples of such population mixing events in Europe are most commonly found in the early Middle Ages. One such example is the Viking migration from Northern Germany to Scandinavia and subsequently to England. In each of these migration steps the migrants mixed with the local population. Due to this mixture, their Germanic DNA signature changed to a Scandinavian-Germanic signature. The more and longer time the Vikings mixed with the locals the faster their German DNA signature decayed. When they moved to England and again mixed with the locals to various degrees, their DNA signature changed again. So, if you are British of Viking descent, the place where your DNA was last mixed may be England or Scandinavia, dependent on your ancestors' history. This is the place that GPS finds. The specific timing of the most recent population mixing event for each individual may of course be different. We conservatively estimate that for most European populations the upper time limit for such events that we can still trace is about 1000 years ago. However, the current version of GPS does not estimate or report the time of said population mixing event.

https://www.prosapiagenetics.com/index.php

My Sister's results.

GPS Results

The GPS tool found your biogeographic point of origin within 787 Kilometeres of Latitude 36.09 and Longitude 46.71.

These results are often seen among individuals of Iranian origin who came from genetic pool similar to yours.

The map below shows the point identified by the GPS along with the points identified by Remix. The map is scalable.

http://i.imgur.com/S7K8dRa.png

http://i.imgur.com/S5qVmqm.png


Kandahari Pashtun

The GPS tool found your biogeographic point of origin within 469 Kilometeres of Latitude 28.41 and Longitude 60.46.

These results are often seen among individuals of Makrani origin who came from genetic pool similar to yours.

The map below shows the point identified by the GPS along with the points identified by Remix. The map is scalable.

http://i.imgur.com/rbajWB5.png

AJL
01-26-2015, 04:25 PM
^Post moved into existing thread on Prosapia/GPS.