PDA

View Full Version : Plot of WHG & S. Eurasian based on K8 Spreadsheet



Kurd
01-15-2015, 04:34 PM
I plotted some K8 spreadsheet populations to aid in visualization of the spreadsheet.
The reason I used WHG and S. Eurasian, is that they are contrasting based on FST (good spread), and that they have a positive North and East geographic correlation, respectively.
Thus, I felt that the plot was informative in that the spatial positioning of the populations on the plot, appears to be consistent with their respective geographic location.
A more complete picture can be formed, if another plot is made using Near Eastern on the y axis, and E. Eurasian or ANE on the x axis, since there appears to be a positive North and East correlation, respectively, with those components as well. The graph may appear too clustered if E. Eurasian isused, due to the relatively low standard deviation in E. Eurasian among many of the K8 populations.
I may be able to add some of the other members who posted their results to the plot(s) over the weekend, time permitting.
Observations: Although some populations seem spatially close toeach other on the plot, they my not be if a 3rd dimension (z axis)is added representing for example E. Eurasian.
The Iranian population which is geographically located between Indiaand the Caucuses, according to the plot shows also to be genetically (based on WHG and S. Eurasian) near Afghan Tajik, Afghan Pashtun, and Afghan Hazara.However, if a 3rd dimension is added to the plot, representing E.Eurasian, the spatial distance to the latter (Hazara) would increase. http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8Spreadsheetlimited_zpscc82f356.jpg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8Spreadsheetlimited_zpscc82f356.jpg. html) Edit: Certain populations were not added, because they caused the plot to become too clustered due to their similarity to other populations

Kurd
01-16-2015, 05:30 AM
Building on the previous post, I plotted the averages of some of the K8 populations. I would like to emphasize that the spatial distances are indicative of genetic distances only for populations that do not have significant amounts of other components, such as NE, E Eurasian, ANE, etc. I will be posting another plot to cover NE and ANE or E Eurasian over the next few days. By viewing both plots side by side (the ones posted so far and the one that I will be posting later), a good idea of genetic distance can be obtained for populations that do not have significant admixture in components outside the ones on both plots (a total of 4 components will be shown on 2 plots). PM me with your results if you would like to be added to the plots.http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8Spreadsheetlimited1_zps3db12e92.jpg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8Spreadsheetlimited1_zps3db12e92.jpg .html)

Krefter
01-16-2015, 05:35 AM
If possible, maybe you should make a single icon for regions. You could also make icons for ethnic groups from the same region the same color. It's hard to keep track what region all the different icons come from.

Kurd
01-16-2015, 05:51 AM
If possible, maybe you should make a single icon for regions. You could also make icons for ethnic groups from the same region the same color. It's hard to keep track what region all the different icons come from.



You could also make icons for ethnic groups from the same region the same color. It's hard to keep track what region all the different icons come from

It will become very difficult to tell one group from the other by looking at the legend, since they will all be the same color.


If possible, maybe you should make a single icon for regions
I'll work on it.

ZephyrousMandaru
01-16-2015, 06:42 AM
Would it be possible for you to add Iraqi and Iranian Jews as well as the two Iraqi Mandaeans? Here are their West Eurasia K8 results.



Iranian_Jewish 14.012975 3.365566667 79.773525 0.082166667 1.019241667 0.240425 0.201941667 1.304166667
Iraqi_Jewish 12.04382308 2.787915385 82.72414615 0.001 0.745584615 0.441638462 0.113223077 1.142676923
IQ_Mandaean 14.13 3.09 80.73 0.00 0.59 0.00 0.00 1.22

Krefter
01-16-2015, 07:01 AM
I would much rather see a graph based on ENF, WHG, and ANE, not South Eurasian, because we don't know what South Eurasian is while the other three are based on ancient genomes. A WHG vs ANE, ANE vs Near eastern, and Near eastern vs WHG for west Asians and north Africans would be interesting. There's a string of WHG in north Africa which may be pre-Neolithic.

Kurd
01-16-2015, 02:52 PM
Would it be possible for you to add Iraqi and Iranian Jews as well as the two Iraqi Mandaeans? Here are their West Eurasia K8 results.



Iranian_Jewish 14.012975 3.365566667 79.773525 0.082166667 1.019241667 0.240425 0.201941667 1.304166667
Iraqi_Jewish 12.04382308 2.787915385 82.72414615 0.001 0.745584615 0.441638462 0.113223077 1.142676923
IQ_Mandaean 14.13 3.09 80.73 0.00 0.59 0.00 0.00 1.22


Sure, no problem

Kurd
01-16-2015, 03:09 PM
I would much rather see a graph based on ENF, WHG, and ANE, not South Eurasian, because we don't know what South Eurasian is while the other three are based on ancient genomes. A WHG vs ANE, ANE vs Near eastern, and Near eastern vs WHG for west Asians and north Africans would be interesting. There's a string of WHG in north Africa which may be pre-Neolithic.

S. Eurasian appears to be an ASE very heavy component, as it appears to peak in the Malay population. While not relevant for European to European comparisons, my intent was to be able do a comparison with S. Asians as well, thus the relevancy.

I do plan to plot a graph of NE vs ANE. With the 2 plots, the one I have posted and the one I will be posting, the following regions' admixture will receive ~ 90 to 99% coverage depending on the population:

1- Europe;
2- W Asia;
3- S Asia, with the exception of E Eurasian heavy populations such as Uzbek, Turkmen, Hazara, etc.

Therefore, to visualize genetic distance, for comparing the aforementioned, you can look at the 2 plots.

The 2 plots will not be very useful for various populations within the following regions, to name some, since they contain significant components of either E Eurasian, SSA, Oceanian, or Pygmy:

1- E Asia
2- N Asia
3- Parts of S Asia and S Central Asia mentioned above
4- Africa
5- Australia

Kurd
01-17-2015, 04:29 PM
Here is the K8 component, ANE vs. NE plot for various populations, including a few Anthrogenica members: http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8SpreadsheetlimitedANE_zps92391235.j pg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8SpreadsheetlimitedANE_zps92391235.j pg.html) EDIT: I will be posting the population means for this plot later. This plot in combination with the previous plot, WHG vs. S. Eurasian can be used to estimate the genetic distance between 2 populations, by adding the linear distance between the two from both plots, subject to the condition that the two populations compared do not have any significant E. Eurasian, Oceanian, Pygmy, and Sub Saharan components.

Krefter
01-17-2015, 04:53 PM
That graph matches PCAs of west Eurasia. It seems the pops that changed west Asia after the early Neolithic had lots of ANE, little to no WHG, and anywhere from 0-70% ENF. The pops that changed Europe after the Neolithic had alot of WHG and ANE, and little ENF.

West Asians look like a mix of something Lezgin-like and Bedouin-like, and Europeans look like a mix of something Stuttgart-like and something Lithuanian-like. The Lithuanian-like pops from bronze age Europe were probably a mix of steppe pops and east and northern hunter gatherers, and spread their genes from east-west and north-south.

The Lezgin pops from Neolithic or bronze age west Asia were probably a mix of something originally from the steppe or south-central Asia, which mixed and became mostly near eastern, and spread from north-south.

It might not be random that an ANE-rich pop from west Asia and an ANE-rich pop from Europe both spread their genes at around the same time. They both could have gotten their ANE from the same Mesolithic source, one mixed with WHG-rich pops and one mixed with ENF-rich pops.

ZephyrousMandaru
01-17-2015, 06:05 PM
The Lezgin pops from Neolithic or bronze age west Asia were probably a mix of something originally from the steppe or south-central Asia, which mixed and became mostly near eastern, and spread from north-south.

I'm of the opinion that Lezgins and other Caucasians were mixed with ANE from the steppe or South-Central Asia, who then interbred with an ENF population from further south (Fertile Crescent) to create the Caucasus component.

ZephyrousMandaru
01-17-2015, 06:12 PM
Nice, my ANE score is typically around 13%, this plot reflects that. Although I can't really see where the Iraqi Jews, Iranian Jews and Mandaeans are. Perhaps we're all so tightly clustered together that that it's difficult to see where they are. Do you think you could do one based on ENF or Near Eastern?

Kurd
01-17-2015, 07:31 PM
Nice, my ANE score is typically around 13%, this plot reflects that. Although I can't really see where the Iraqi Jews, Iranian Jews and Mandaeans are. Perhaps we're all so tightly clustered together that that it's difficult to see where they are. Do you think you could do one based on ENF or Near Eastern?

You were all practically on top of each other. You can see them by trying to zoom in. The only way around it was for me to limit the ANE range on the plot to say 25% and the NE range to about 70%, but that would have meant that some populations would have fallen off the chart.

The other way around the problem is to increase the resolution on the graphic. That however, is also not possible, since we are limited on our file uploads to 100Kb on Anthrogenica.

When you say based on ENF, which spreadsheet or calculator are you referring to?

EDIT: I can probably upload a higher resolution graphic to a photo sharing website, and provide a link in the post. Interested members can print the plot from the higher resolution graphic.

Kurd
01-18-2015, 12:54 AM
Here is the plot for the K8 ANE vs. Near Eastern Population Means, with a few Anthrogenica members added:

http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8SpreadsheetlimitedPLOTANE_zpsc7c5d3 61.jpg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8SpreadsheetlimitedPLOTANE_zpsc7c5d3 61.jpg.html)

PM me if you would like to be added to any future plots.

Chad Rohlfsen
01-18-2015, 01:11 AM
Nice! I think that I would be right above Krefter.

ZephyrousMandaru
01-18-2015, 01:15 AM
You were all practically on top of each other. You can see them by trying to zoom in. The only way around it was for me to limit the ANE range on the plot to say 25% and the NE range to about 70%, but that would have meant that some populations would have fallen off the chart.

The other way around the problem is to increase the resolution on the graphic. That however, is also not possible, since we are limited on our file uploads to 100Kb on Anthrogenica.

When you say based on ENF, which spreadsheet or calculator are you referring to?

EDIT: I can probably upload a higher resolution graphic to a photo sharing website, and provide a link in the post. Interested members can print the plot from the higher resolution graphic.

Yeah, I figured we all were. You could try uploading the photographs to TinyPic or another image hosting website and posting the image that way. By ENF I was referring to Near Eastern from West Eurasia K8, ENF was the term used for Near Eastern with David's Eurogenes ANE K=7.

Chad Rohlfsen
01-18-2015, 01:54 AM
I posted a tweet by Razib Khan earlier. There is a ton of great papers coming out, hopefully soon. One indicates that Jomon could be basal to East Eurasian. David has posted the abstract from the site. It looks like a good one!!

http://sarkoboros.com/2015/01/16/oist-ancient-dna-symposium/

Have a look!

Kurd
01-18-2015, 06:15 PM
http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8SpreadsheetlimitedANEVSNE_zps89bc0f 10.jpg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8SpreadsheetlimitedANEVSNE_zps89bc0f 10.jpg.html)

Assuming the underlying assumptions to calibrate the calculator are good, which they do appear to be, a couple of observations can be made based on this plot/K8 spreadsheet, and the one previously posted S Eurasian vs WHG). Perhaps members more knowledgeable than myself (DMXX, Parasar, .......) can weigh in:

1- Neolithic Europeans during the time of Stuttgart (~70%NE) and Loschbour (0 NE) lie around the y-axis, and can be modeled as a mixture of the two with the Stuttgart component decreasing northward.

2- Present Europeans have been ANE shifted from the past, and can be modeled on the A-axis as a mixture of ENF/Bedouin and Proto N. European (WHG + ANE). Groups such as the S. Italians having a much higher percentage of the former, than N. Europeans.

3- Unlike Europeans, Iranic people (B-axis) can better be modeled as a mixture of ENF and Proto S.or SC Asian, or alternatively, as a mixture of Caucus and Proto S. or SC Asian.

4- There appears to be a genetic gap on the ANE plot between Iran and the Indus Valley groups. I wonder if this is a result of some isolation caused by physical barriers such as Iranian deserts, that caused the Indus Valley populations to become relatively isolated from Iranian populations.

parasar
01-18-2015, 07:39 PM
...
1- Neolithic Europeans during the time of Stuttgart (~70%NE) and Loschbour (0 NE) lie around the y-axis, and can be modeled as a mixture of the two with the Stuttgart component decreasing northward.

2- Present Europeans have been ANE shifted from the past, and can be modeled on the A-axis as a mixture of ENF/Bedouin and Proto N. European (WHG + ANE). Groups such as the S. Italians having a much higher percentage of the former, than N. Europeans.

3- Unlike Europeans, Iranic people (B-axis) can better be modeled as a mixture of ENF and Proto S.or SC Asian, or alternatively, as a mixture of Caucus and Proto S. or SC Asian.

4- There appears to be a genetic gap on the ANE plot between Iran and the Indus Valley groups. I wonder if this is a result of some isolation caused by physical barriers such as Iranian deserts, that caused the Indus Valley populations to become relatively isolated from Iranian populations.


I too think West to East it the desert that causes the discontinuity. The populations to look at would be those from Sistan-Baluchistan in Iran. If they are found to be closer to Balochi-Makarani-Brahui than the Kermani, then a discontinuity due to the Lut desert wound be confirmed.

Chad Rohlfsen
01-18-2015, 07:52 PM
That gap is the 8-20% gap in ANE between Iranians and SC Asian groups.

Kurd
01-18-2015, 09:06 PM
I too think West to East it the desert that causes the discontinuity. The populations to look at would be those from Sistan-Baluchistan in Iran. If they are found to be closer to Balochi-Makarani-Brahui than the Kermani, then a discontinuity due to the Lut desert wound be confirmed.

We have a new member whom I believe may be from Irani Baluchistan (zarafshaan). He may have access to other DNA results from his area, or perhaps even the Kerman area.

Kurd
01-18-2015, 10:10 PM
I cleaned it up and added a few additional members:

http://i1377.photobucket.com/albums/ah42/dil88888/West_Eurasia_K8SpreadsheetlimitedWHG_zpse909661d.j pg (http://s1377.photobucket.com/user/dil88888/media/West_Eurasia_K8SpreadsheetlimitedWHG_zpse909661d.j pg.html)

The Barnacle
01-19-2015, 02:38 PM
That gap is the 8-20% gap in ANE between Iranians and SC Asian groups.

South Central Asian are afghans mainly. Iranian typically score around 19-21% ANE, afghans typically score 28-31% ANE so it's around 9% difference I beleive not 20%

Chad Rohlfsen
01-19-2015, 03:18 PM
Close to 20% when you get to the Kalash.

Chad Rohlfsen
01-20-2015, 02:37 AM
Here is a bit of a teaser on the Jomon study. It is by the same author, from 2013. Maybe it was part of his doctorate study and just now developed into a full study. Anyway, it includes aDNA, fst, d-stats, and a whole lot of great stuff!!!! The best part starts at p124, but don't miss out on the rest. Don't worry about the link. It's a PDF. Enjoy!

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=5&ved=0CDYQFjAE&url=http%3A%2F%2Fir.soken.ac.jp%2F%3Faction%3Drepo sitory_uri%26item_id%3D4960%26file_id%3D19%26file_ no%3D2&ei=CIO9VLeFO4OYyASv7oCgAw&usg=AFQjCNE9l1_7Z3jO3v28v_UJztOt2YVYTw&sig2=B6gIGkRtkOeHfNNSOazbPA

parasar
01-20-2015, 03:06 PM
That gap is the 8-20% gap in ANE between Iranians and SC Asian groups.


ANE may be responsible for causing that gap, but the gap is in the W. Eurasian Basal cline with Sardinians at one end and southern Indians at the other. Perhaps an intrusion of ANE in the middle from near the northern Indus causes the break (cf. Burusho, Kalash, Pathan).
On the Y-DNA side we also see that within Iran east of the Lut desert, Sistan-Baluchistan clusters with the Indus populations - "five main clusters: East-African, North-African and Near Eastern Arab, European, Near Eastern and South Asian ... Sistan Baluchestan (SB-Ba) that clusters with its neighbouring Pakistan."
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0041252
http://2.bp.blogspot.com/-eW1ewu33_eA/UAgkXmyyyQI/AAAAAAAAFDM/GDG06jc222Q/s1600/journal.pone.0041252.g003.png

Chad Rohlfsen
01-21-2015, 09:25 PM
What is Salkin's ancestry?

randwulf
01-21-2015, 09:49 PM
What is Salkin's ancestry?

In another post he said he was Swedish with a small amount of Belgian.

Chad Rohlfsen
01-22-2015, 02:21 AM
In another post he said he was Swedish with a small amount of Belgian.

Okay, thanks! I was curious since he has a little more WHG than myself.