PDA

View Full Version : Odd Geno 2.0 Results for McElrea



David Mc
09-12-2013, 07:10 AM
I received my Geno 2.0 results today. The Paternal and Maternal results duplicated FTDNA's (except, of course Geno 2.0 doesn't refine L21 down to DF49), but the "Who Am I" results seemed a bit odd. They show me as 42% Northern European, 37% Mediterranean, and 19% Southwest Asian.

Using the above breakdown, they give my first reference population as German (which is relatively close) and my second reference population as Greek (which is not so close). Here's the thing. My father is Northern Irish-- and all of his ancestors are Northern Irish-- as far back as our records go (1750 AD). The jury is out as to whether or not his paternal line is Scottish or Native Irish, but most of the autosomal DNA would be of Scots derivation. My mother is Native Irish/English, again, as far back as the records go. How do I end up being matched to German reference groups and (to a lesser extent) Greek reference groups?

I appreciate any insight from those who have had more time to ponder the Genographic Project's work.

David McElrea

David Mc
09-17-2013, 06:30 AM
A bump with screenshots and further questions:

Here is the main panel from the "Who Am I" section:

681

I was expecting to fall neatly within the British reference population. All of my ancestors are from the Isles, both recent and immediate. As it happens, though, I have significantly less Northern European admixture than the average Briton, and more Mediterranean and Southwest Asian.

My first reference population is German:

682

The German-like admixture was surprising enough, given my ancestry, but not nearly as surprising as my y second reference group, which is (rather oddly, as there are significant differences in admixture percentages), Greek:

683

I don't know how meaningful the population references are, but they obviously signify something. This is where I would appreciate some help from my fellow-forumites. I don't know what they signify.

My first question, then, is simply that: what do they signify? Does the elevated Mediterranean admixture suggest that one of my recent ancestors (say, within the last two to three hundred years) came to Britain from the Continent?

Secondly, have other British Isles testees come up with similar variances from the norm in their percentages?

These first questions have to do with my own results. My last question addresses the larger question of Geno 2.0's descriptors.

Geno 2.0 describe Northern European as reflecting Europe's earliest inhabitants. This, as far as I understand, is fairly uncontroversial. The Mediterranean and Southwest Asian designations are both tied in with neolithic incomers from the Near East. Am I right in saying this is contested ground for many on the forum, at least with regard to the Southwest Asian?

Any help would be appreciated. I'm far more comfortable talking about SNP's than I am autosomal DNA, and I'm feeling quite out of my depth in figuring out what all of this means with regards to ancestral populations and my own autosomal make-up.

Scarlet Ibis
09-17-2013, 11:00 PM
My first question, then, is simply that: what do they signify? Does the elevated Mediterranean admixture suggest that one of my recent ancestors (say, within the last two to three hundred years) came to Britain from the Continent?

Secondly, have other British Isles testees come up with similar variances from the norm in their percentages?


I hate to give such an inconclusive answer, but: maybe, and maybe not. Your results look within the norm for the few British results I've seen.

For example, your results look identical to Solothurn's, and he's 100% English, as far as I know. http://www.anthrogenica.com/showthread.php?36-R1b-U-152-Updates-News&p=7900&viewfull=1#post7900

I also believe Quotablepatella, who is also 100% English, got the same reference population results, and 37% Mediterranean. I'll ask him if he can post his results here.

AJL
09-18-2013, 01:00 AM
I don't think Geno 2 is really that great at population analysis -- certainly not as good from what I've seen as 23andme's Countries of Ancestry/Ancestry Composition in tandem.

David Mc
09-18-2013, 02:04 AM
Thank you both for responding. I was more than a little bemused.

Scarlet Ibis
09-18-2013, 03:26 AM
I don't think Geno 2 is really that great at population analysis -- certainly not as good from what I've seen as 23andme's Countries of Ancestry/Ancestry Composition in tandem.

Yeah, from what I've seen, Geno's population analysis leaves something to be desired. They're fun to look at, and they're not horrible like DNATribes is (sorry, but anything that charges money to tell Brits they're "Indian" is seriously horrible), but I wouldn't put too much stock in them. 23andme is a much more useful tool in this regard, particularly since you can use the raw data to play around with Gedmatch, and the various MDLP, Dienekes, Eurogenes, etc. DIY calculators.

A late welcome to Anthrogenica, btw, David Mc. :welcome:

David Mc
09-18-2013, 04:20 AM
Thanks very much, Scarlet Ibis.

dallitt
06-02-2014, 09:47 AM
The same David. I got German first and Greek second. I have a similar ancestry. I thought it might be because Irish people's ancestry is often Basque and they might be the Indo-European connection. Other possibility is Geno 2 doesn't have it's act together and the results are too broad.

David Mc
06-02-2014, 06:52 PM
Hi Dallitt. I'm thinking the latter is the most likely explanation. I'm afraid I "pulled the trigger" a little too quickly with Geno 2.0... it hasn't been at all what I hoped it would be.

David Mc
06-02-2014, 09:02 PM
Thought I should make two additional comments to elucidate the previous post. Firstly, there has been a significant shift away from the Iberian origins model for the Irish. I won't present all of the reasons for the shift here as there are numerous threads on this forum devoted precisely to that topic. As I don't hold to there being a significant Iberian input to the Irish gene pool, I think it more likely that the problem lies in the way the Genographic Project gathers or collates (or perhaps defines) their data.

Secondly, I thought I should say why I've been disappointed with my Geno 2.0 experience, rather than simply lobbing a verbal hand grenade at the Project without explanation. When I signed up for the test I knew that my own DF49 subclade was not included on the new chip, so I can't complain about that. I had hoped that it would at least offer up the possibility of new SNP's below DF49, but none of that has materialized, despite all the hype. Furthermore, the access customers have to Geno 2.0 data (matches etc) is very limited and/or imprecise. Apart from identifying my maternal haplogroup (which I could have done elsewhere for less money), I've learned nothing new from my Geno 2.0 experiment. That is why I'm disappointed. Caveat Emptor, as they say.

dp
10-31-2014, 03:32 PM
Dear David,
I'm sorry to read your disappointment on your Geno 2 test. At least your Big-Y test can give you novel results. dp :-)
PS: AJL (http://www.anthrogenica.com/showthread.php?1317-Odd-Geno-2-0-Results-for-McElrea&p=13908&viewfull=1#post13908) and Scarlet Ibis (http://www.anthrogenica.com/showthread.php?1317-Odd-Geno-2-0-Results-for-McElrea&p=13915&viewfull=1#post13915). Your posts make me look forward to seeing what 23andme will say about my aunt's ethnicity. I also like that I can put them on GEDmatch and compare them to the kits I've tested with FamilyTreeDNA.
dp :-)

JohnOBrien
01-04-2015, 04:11 PM
Hi Dallitt. I'm thinking the latter is the most likely explanation. I'm afraid I "pulled the trigger" a little too quickly with Geno 2.0... it hasn't been at all what I hoped it would be.

You are not the Lone Ranger, David, my GENO 2.0 results told me my first reference population is "British (United Kingdom)," and my second reference population is "German." Well, Woo Hoo, I knew I was a Kraut-Mick ever since I was a wee little boy. I transferred to FTDNA and ordered both the 37 and 67 marker tests. Results are slow to arrive. I must post that at the present time my expectations are not very high. Oh, I did order the FTDNA "Family Finder" also. We shall see what happens. I will post the results.

vettor
01-04-2015, 05:43 PM
I received my Geno 2.0 results today. The Paternal and Maternal results duplicated FTDNA's (except, of course Geno 2.0 doesn't refine L21 down to DF49), but the "Who Am I" results seemed a bit odd. They show me as 42% Northern European, 37% Mediterranean, and 19% Southwest Asian.

Using the above breakdown, they give my first reference population as German (which is relatively close) and my second reference population as Greek (which is not so close). Here's the thing. My father is Northern Irish-- and all of his ancestors are Northern Irish-- as far back as our records go (1750 AD). The jury is out as to whether or not his paternal line is Scottish or Native Irish, but most of the autosomal DNA would be of Scots derivation. My mother is Native Irish/English, again, as far back as the records go. How do I end up being matched to German reference groups and (to a lesser extent) Greek reference groups?

I appreciate any insight from those who have had more time to ponder the Genographic Project's work.

David McElrea

Download your numbers from Natgeno ( you should have over 14000 lines on the spreadsheet).........then go to Felix site and run the program that will convert those numbers into positive and negative SNP ...................

BTW when transferring to Ftdna, they should accept ALL positive and negative SNP's, but alas they do not.................so you are left with extra positive and negative SNP found via felix program for you natgeno data that "does not" fit ftdna tree ..........................I never understood why ftdna does not accept ALL of natgeno2 data especially since they work together

wombatofthenorth
03-29-2015, 11:17 PM
Yeah, from what I've seen, Geno's population analysis leaves something to be desired. They're fun to look at, and they're not horrible like DNATribes is (sorry, but anything that charges money to tell Brits they're "Indian" is seriously horrible), but I wouldn't put too much stock in them. 23andme is a much more useful tool in this regard, particularly since you can use the raw data to play around with Gedmatch, and the various MDLP, Dienekes, Eurogenes, etc. DIY calculators.

A late welcome to Anthrogenica, btw, David Mc. :welcome:

I could be wrong but I think 23andme tries to focus mostly on going only 6 generations back so they can try to get it down to various tiny sub-regions and countries while Geno 2.0 admixture looks deeper and is a thus a mix of prior 6 generations plus semi-deep ancestry and just gives percentages from the nine basic large scale semi-old types, in Europe many countries and regions might share very similar percentages.

wombatofthenorth
03-29-2015, 11:21 PM
I didn't get a chance to look it over much yet, but apparently they just came out with a hyper-detailed study of the UK, apparently a few sections have much different admixture than the bulk, perhaps something here might explain it. I forget which regions they said were different. I think one region has way more Northern Europe and one considerably less than the rest. The article was in the New York Times a couple days ago (and elsewhere). Maybe your results would actually match the one with lower N.Europe closely if Geno 2.0 had a specific reference population for that???

wombatofthenorth
03-29-2015, 11:23 PM
Download your numbers from Natgeno ( you should have over 14000 lines on the spreadsheet).........then go to Felix site and run the program that will convert those numbers into positive and negative SNP ...................

BTW when transferring to Ftdna, they should accept ALL positive and negative SNP's, but alas they do not.................so you are left with extra positive and negative SNP found via felix program for you natgeno data that "does not" fit ftdna tree ..........................I never understood why ftdna does not accept ALL of natgeno2 data especially since they work together

How do you download from Nat Geo? Did they remove the link since your post?

Huntergatherer1066
03-29-2015, 11:30 PM
How do you download from Nat Geo? Did they remove the link since your post?

you should be able to click on the little person icon in the upper right corner and click on my settings. that'll take you to a page where there is an option to "Download Genetic Data".

vettor
03-29-2015, 11:41 PM
How do you download from Nat Geo? Did they remove the link since your post?

when you log in, go to your name ( top right corner) and select myprofile. then go down to bottom of page, where it says download genetic marker.
You will end up with a 14000plus lines on a spreadsheet.

you can then go to felix program ( if you like ) and it coverts the RS#'s into SNP's

since natgeno tests every marker , the non positives SNP are all negatives and does not refer to as untested SNP's

wombatofthenorth
03-29-2015, 11:54 PM
you should be able to click on the little person icon in the upper right corner and click on my settings. that'll take you to a page where there is an option to "Download Genetic Data".

Ah, thanks, didn't realize it was hidden away there.

David Mc
03-30-2015, 12:11 AM
I didn't get a chance to look it over much yet, but apparently they just came out with a hyper-detailed study of the UK, apparently a few sections have much different admixture than the bulk, perhaps something here might explain it. I forget which regions they said were different. I think one region has way more Northern Europe and one considerably less than the rest. The article was in the New York Times a couple days ago (and elsewhere). Maybe your results would actually match the one with lower N.Europe closely if Geno 2.0 had a specific reference population for that???

I think you're probably thinking of the "People of the British Isles" study that was just released. It certainly would be interesting to look at the actual data for the different population groups, but (alas) they are not going to be publishing the test-results.

Since first writing the original post, I've come to the conclusion that Geno 2.0 just isn't a reliable guide for locating individual samples on the admixture spectrum. I'd love to know the parameters they use, because they seem to differ wildly from the results one finds on GEDmatch. Granted, one can get pretty dissimilar results from GEDmatch's various admixture tools, as well... but hey, this is still an imprecise science.

wombatofthenorth
03-30-2015, 03:15 AM
I think you're probably thinking of the "People of the British Isles" study that was just released. It certainly would be interesting to look at the actual data for the different population groups, but (alas) they are not going to be publishing the test-results.

Since first writing the original post, I've come to the conclusion that Geno 2.0 just isn't a reliable guide for locating individual samples on the admixture spectrum. I'd love to know the parameters they use, because they seem to differ wildly from the results one finds on GEDmatch. Granted, one can get pretty dissimilar results from GEDmatch's various admixture tools, as well... but hey, this is still an imprecise science.

Yes.

Interestingly, looking at one of the plot charts associated with the study it appears to show that the North Irish section actually does have more input from Spain and France than the bulk of England, assuming I am even reading it all correctly. Perhaps the English reference population for Geno 2.0 was mostly derived from the main bulk English population and thus your North Irish doesn't match as well, coming up with too much Mediterranean?

OTOH, I thought it was said that the differences were supposed to be generally slight and that they had to use special tools, so would these component differences really be enough to flip the Geno 2.0 results you got from to 49NE/33M to 42NE/37M?? You seem to have an extra 1% missing compared to most so even if we toss that to NE it's still 49 vs 43 and 33 vs 37? Maybe it's possible though?

I suppose perhaps you've already seen the chart anyway by now.

Yeah it is quite a shame the data is basically locked away from most.