PDA

View Full Version : New myOrigins Results Made Me Do This: DNA.Land



curiousII
04-11-2017, 04:41 AM
I've posted my old myOrigins map in a couple of threads around here but here's the results again anyway:

British Isles: 59%
Eastern Europe: 13%
Western/Central Europe: 11%
Finland/Northern Siberia: 9%
Scandinavia: 6%
Middle Eastern: 2%

Here's the new myOrigins:

British Isles: 13%
Eastern Europe: 5%
Western/Central Europe: 56%
Scandinavia: 20%
Iberia: 4%

Both test results gave me 98% European. The missing 2% is attributed to possible "statistical noise." But, like so many other persons (judging by the posts you read on the FTDNA Forums site) I got sort of flummoxed by the switch. So, uncaring of just how much worse things could get, I tried DNA.Land. Here's how much I regret that now:

West Eurasian 100%
Northwest European 64%
Northeast European 20%
North Slavic 17%
Finnish 2.5%
Southwestern European 8.9%
Ashkenazi 3.8%
Balkan 2.5%
Ambiguous 1.2%

So, anyway, this is me according to them. Going by the DNA.Land home page it only had 44,815 members. Any idea how many DNA tests FTDNA has in its database to compare with? I know it just made a great adjustment to its populations/calculators/however it is they do all that, and I'm certain it has to be more than DNA.Land's 44,000. Will DNA.Land adjust its results as it goes along and obtains more tests? Or is this a typical comparison between the two companies?

The Ashkenazi hit: I believe this is the first time I've seen that in any tests I've taken. The FTDNA myOrigins does give me <2% Ashkenazi statistical noise in its new results, but with DNA.Land is this an appreciable amount? I've read on other sites that if a person has any Jewish results he would also have SSA. But it doesn't appear that I have any measurable SSA; does that mean the two don't necessarily go together and that makes this also statistical noise? Ashkenazi doesn't match any of my known family history, that's next logical question and answer. And I think that's a good question and this is a good time to ask it, what with all the new adjustments to the DNA calculators: Is Jewish ancestry synonymous with SSA?

Does Ashkenazi show up regularly in Eastern European/Siberian/Finn hits? I've read both yes and no if I'm remembering correctly, and we all know now just how drastically a company can change its customers' test results. I suppose that's the fun part of taking these tests is anticipating the shock when you get the results. And it's not just once, it's whenever FTDNA and its contemporaries decide to adjust their chips or whatever it is they use.

Like they say: It's all good.

geebee
04-11-2017, 05:43 AM
I find it interesting that your British Isles went down and your Western/Central Europe went up. My impression had been that more people experienced the opposite.

Of course, that's entirely subjective on my part and may not be true at all. Here's what I can tell you about my family, though.

My brother Bernie. British Isles went from 56% to 81%. He didn't have any Western/Central Europe, though I'd have expected him to, but he did have 22% Southern Europe and 15% Scandinavia. Both of these seem to have disappeared, though he did end up with 4% Southeast Europe. His 1% Middle Eastern (North Africa) disappeared, but 8% Middle Eastern (Asia Minor) appeared instead. Also, his total Europe dropped from 93% to only 85%. If you add the Middle Eastern, you get back almost to the previous combined European/Middle Eastern total. However, the decision to report anything below 2% as trace (making no distinction between something that's 0.1% and something that's 1.9%) is certainly annoying.

My brother Curt. Only a modest increase in British Isles, from 44% to 51%. Scandinavia 33% disappeared, but West and Central Europe went from 6% to 31%. Southern Europe 11% disappeared; Southeast Europe 3% appeared. East Europe 10% also appeared. Remainder of results are "trace".

Me. British Isles went from 50% to 70%; Southern Europe 30% vanished; Scandinavia 13% vanished. Middle Eastern (Eastern Middle East) 2% replaced by Middle Eastern (Asia Minor) 13%. Remainder of results are "trace".

My sister Kim. Useless 100% Southern, Western, and Central Europe replaced by 60% British Isles, 23% Southeast Europe, 10% Scandinavia, and 4% Iberia; plus trace results.

My sister Luci. British Isles 58% stayed nearly the same, dropping 1%. Southern Europe 16% disappeared. Scandinavia increased from 15% to 18%. Jewish Diaspora dropped 1%, from 6% to 5%. Middle Eastern (East Middle East) decreased from 5% to trace. Remainder of results are "trace".

My sister CJ. British Isles increased from 59% to 83%; Scandinavia 18% vanished; Southern Europe 15% vanished. Southeast Europe 14% appeared. Middle Eastern 4% (Asia Minor 1%, Eastern Middle East 1%, North Africa 1%) vanished. Remainder of results are "trace".

My daughter Kathryn. Ridiculous 97% Southern, Western, and Central Europe and 3% Finland and Northern Siberia transformed into 89% British Isles, 5% Jewish Diaspora (Ashkenazi), 5% Middle East (Asia Minor), plus trace results.

Kathryn's mother Diane. 87% British Isles increased to 94% British Isles. 5% Southern Europe and 1% Eastern Europe disappeared. 5% Middle Eastern (Asia Minor) increased to 6% Middle Eastern (Asia Minor). Remainder of results are "trace".

I don't think I'm being too harsh in labeling Kathryn's previous results as ridiculous. Both of her parents had much more specific results, so I don't see why Kathryn shouldn't have. As to her current results, her British Isles is a bit higher than her mother's -- which is even higher than mine -- but not unreasonably so. Given the total for both parents, Kathryn's amount would certainly not be unreasonable. What's kind of odd, though, is that both Diane's and my Middle East is labeled solely as "Asia Minor".

So why does Kathryn show both 5% Middle East (Asia Minor) and 5% Jewish Diaspora (Ashkenazi). If neither parent has Jewish Diaspora, then neither can Kathryn.

EDIT: I can definitely see that even among my family members, British Isles didn't go up universally. But I don't think any of us had any significant loss of British Isles in favor of Central/West Europe.

The odd thing is that even my dad, whose results I didn't include, fails to show any Central/West Europe in MyOrigins 2.0, despite being about half-and-half German and British.

His old results were a vague 100% Southern, Western, and Central Europe. The current results are 76% British Isles, 15% Southeast Europe, 8% East Europe, plus "trace" Northeast Asia.

I think that in my dad's case -- and that of a lot of people -- his Palatine German ancestry is somehow split between British Isles and Southeast Europe/East Europe. Much of his ancestry may have been from southern Germany, they were still from Germany -- or at least, part of what became Germany -- as far as we can tell.

curiousII
04-11-2017, 07:28 AM
I find it interesting that your British Isles went down and your Western/Central Europe went up. My impression had been that more people experienced the opposite.

And that's what I've noticed from reading the Forum posts, too. Everyone's more UK now: http://forums.familytreedna.com/showthread.php?t=41497

Except me, anyway. As you sort through those posts you'll see many customer's UK has increased exponentially. Mine dropped and was added to West/Central Europe, along with a new Iberia (I see you list Catalan. I don't think I'm Basque, though I am DF27>Z2573. That's said to be from the Pyrenees).

I'm also Western Atlantic Modal Haplotype (WAMH). I just checked my FTDNA match page: I have 1,925 FF matches, 5,690 y-DNA 12 Marker matches, and I have 251 mtDNA H11a matches. I mention this as I'm pretty common in the New World according to FTDNA; I have no matches at all with DNA.Land. I double-checked there, still "Unfortunately, we do not have any matches for you in our database." And they've even gained a couple of new genomes since my last post, but still no hits for me.

Is this an example of something? Different kinds of customers between the two companies? Different types of calculations used due to this? Now this poster's example of how much his sibling's FF result and map differ from his even though they the same biological parents is a possible example of something similar to your question concerning your child Kathryn's anomaly:

http://www.anthrogenica.com/attachment.php?attachmentid=11975

He speaks of it here, important so I'm not taking his family's genetics out of context. Post #456: http://www.anthrogenica.com/showthread.php?827-Where-did-DF27-originate-and-when-and-how-did-it-expand&p=224946#post224946

Anyhow, that's what DNA.Land just did to me. All three tests, FTDNA's myOrigins #1 and #2, coupled with DNA.Land's fabulous interpretation of my internals, have given me a lot of new conversation. That's fun, it's all good. And, thank you for sharing the ambiguous results of your family's. This next is detailing the change in an acquaintance's myOrigins which I purchased for her as a gift along with an mtDNA test. You'll easily see just how much her results have changed, and since she's Swiss I'd really like to upload her to GEDMatch and DNA.Land for comparison, but I won't as I've yet to get her permission for that. Her old myOrigins with 99% European:

56% Southern Europe
39% Scandinavia
4% British Isles followed by 1% Central Asia.

New, still 99% European but the missing 1% is now unaccounted for:

48% West/Central Europe
25% Southeast Europe
16% British Isles
10% Scandinavia

See that originally she had no West/Central Europe; now that region's where she maxed out.

An edit: I found this, it's about a year old: https://medium.com/@dl1dl1/the-new-ancestry-report-f2844384259b

I've read in a couple different places now about DNA.Land using a "different algorithm." OK, going by the example given of Mr. Erlich's results you'll see he has a West Eurasian result of 100% even though he's got Arab/Egyptian 5.3%. Could that be part of an Indo-Iranian immigration event into Europe in the deep past? My guess is no as this population is listed separately from the Indo-Iranians.

So how could Arab/Egyptian be considered West Eurasian?

Finn
04-11-2017, 04:56 PM
I've posted my old myOrigins map in a couple of threads around here but here's the results again anyway:

British Isles: 59%
Eastern Europe: 13%
Western/Central Europe: 11%
Finland/Northern Siberia: 9%
Scandinavia: 6%
Middle Eastern: 2%

Here's the new myOrigins:

British Isles: 13%
Eastern Europe: 5%
Western/Central Europe: 56%
Scandinavia: 20%
Iberia: 4%

Both test results gave me 98% European. The missing 2% is attributed to possible "statistical noise." But, like so many other persons (judging by the posts you read on the FTDNA Forums site) I got sort of flummoxed by the switch. So, uncaring of just how much worse things could get, I tried DNA.Land. Here's how much I regret that now:

West Eurasian 100%
Northwest European 64%
Northeast European 20%
North Slavic 17%
Finnish 2.5%
Southwestern European 8.9%
Ashkenazi 3.8%
Balkan 2.5%
Ambiguous 1.2%

So, anyway, this is me according to them. Going by the DNA.Land home page it only had 44,815 members. Any idea how many DNA tests FTDNA has in its database to compare with? I know it just made a great adjustment to its populations/calculators/however it is they do all that, and I'm certain it has to be more than DNA.Land's 44,000. Will DNA.Land adjust its results as it goes along and obtains more tests? Or is this a typical comparison between the two companies?

The Ashkenazi hit: I believe this is the first time I've seen that in any tests I've taken. The FTDNA myOrigins does give me <2% Ashkenazi statistical noise in its new results, but with DNA.Land is this an appreciable amount? I've read on other sites that if a person has any Jewish results he would also have SSA. But it doesn't appear that I have any measurable SSA; does that mean the two don't necessarily go together and that makes this also statistical noise? Ashkenazi doesn't match any of my known family history, that's next logical question and answer. And I think that's a good question and this is a good time to ask it, what with all the new adjustments to the DNA calculators: Is Jewish ancestry synonymous with SSA?

Does Ashkenazi show up regularly in Eastern European/Siberian/Finn hits? I've read both yes and no if I'm remembering correctly, and we all know now just how drastically a company can change its customers' test results. I suppose that's the fun part of taking these tests is anticipating the shock when you get the results. And it's not just once, it's whenever FTDNA and its contemporaries decide to adjust their chips or whatever it is they use.

Like they say: It's all good.

For me form North Dutch stock DNA Land is more spot on than My Origins 2.

My Origins 2.0. :
http://i65.tinypic.com/2rxbtir.jpg

DNA Land:
http://i68.tinypic.com/1zd40gl.jpg

I've had contact with FTDNA they admit that a NW European admixture is missing in my Origins 2.0.

curiousII
04-12-2017, 04:41 AM
I just found this: http://forums.familytreedna.com/showthread.php?t=38816

MMadi's Post #3: "Unless they've fixed their algorithm, dna.land is giving too much Ashkenazi to people, especially those with Southern European (Italian, Greek) ancestry." OK, this was posted in November, 2015, which was maybe six months before DNA.Land switched its algorithm or whatever it did. But see the Ashkenazi results were drawing questions and criticism close to two years ago?

But congrats on your DNA.Land result, Finn. Clean and concise.

Calas
04-12-2017, 09:24 AM
I just found this: http://forums.familytreedna.com/showthread.php?t=38816

MMadi's Post #3: "Unless they've fixed their algorithm, dna.land is giving too much Ashkenazi to people, especially those with Southern European (Italian, Greek) ancestry." OK, this was posted in November, 2015, which was maybe six months before DNA.Land switched its algorithm or whatever it did. But see the Ashkenazi results were drawing questions and criticism close to two years ago?

But congrats on your DNA.Land result, Finn. Clean and concise.

DNALand still hasn't fixed that Ashkenazi issue. It is also a lesser problem with 23&me given the way that their algorithm is arranged. Both mistake certain Southern European genes with Ashkenazi. It likely happens with all the ethnicity genetic tests to some degree hence the ever so common "I didn't know I had Ashkenazi" on various forums.

There was a case study a while back about that and a similar ancestry algorithm as 23&me where a guy was "20% Ashkenazi". Shame when compared against legitimate Ashkenazi he (and likely others) was about as legit as a banana being passed off as a blueberry.

galon07
04-12-2017, 01:28 PM
In my case, myOrigins 2.0 and DNA.Land actually showed more similarities than differences. Here's for comparison:

myOrigins 2.0:

62% European
29% Western and Central Europe
19% Southeast Europe
14% Iberia

12% New World
9% North and Central America
3% South American

11% Middle Eastern (all North Africa)

9% African (all West Africa)

Trace results for Finland, Siberia, Oceania, and West Middle East


DNA.Land:
68% West Eurasian
34% South/Central European
24% Southwestern European
7.6% Finnish
2.4% Northwest European

17% African
9.6% North African
5.1% Lower Niger Valley
1.9% East African

13% Native American
10% Native American
3.2% Amazonian

1.8% Ambiguous

I'm not sure about the Ashkenazi problem you mentioned, but none of the companies showed any trace of the alleged Jewish ancestry coming from my father's mother side. I saw here that many people (just like you) unexpectedly got this result, but for me it was just the opposite.

curiousII
04-12-2017, 02:37 PM
In my case, myOrigins 2.0 and DNA.Land actually showed more similarities than differences. Here's for comparison

That is quite consistent. That's sort of a rarity for competitors to provide their customers similar results, from what I've read. And that's a question I haven't found an answer to yet from browsing these DNA sites: Why would different companies use different population models (or whatever they're called) when they set their algorithms? History pretty much agrees on past world migrations routes, dates, and peoples/cultures now, doesn't it? History classes in varying countries are fairly much in agreement with that part of world history, aren't they?

Or is there still enough disputes and bickering about ethnic origins to cause different DNA companies to use algorithms to calculate a customer's ethnic identity and ancestral history? This is a new science and change is good and all, but these companies have been established for a considerable time now and it would seem that they would have refined their research to reach even a partial agreement on the planet's ethnic groupings.

Or I'm entirely wrong. I've read that it's illegal in France to partake in DNA testing, and a large part of Iberia is untested due to the population's financial plight. The latter meaning that they'd rather spend their money on food and essentials rather than DNA tests. I understand the problem there and I suppose there's similar difficulties in other parts of the globe. But with as much professionalism and education that these DNA company's founders and employees have, you'd think they'd have surmounted much of this by now. I mean, DNA.Land's hooked up with Columbia University, a college that's generally acknowledged as being far from a diploma mill.


I'm not sure about the Ashkenazi problem you mentioned

No, it's not a problem. I seem to be following suit with other customers who have a surprise heritage pop up in their results.

I just took the DNA.Land Trait Prediction Report. You know, they don't have my eye color in their examples. I drink more coffee than the average DNA.Land customer; if they'd included energy drinks I'd be far off the scale. RockStar has a new Black Cherry with 240 mg. of caffeine that tastes great.

galon07
04-13-2017, 12:23 PM
Yes, I think the main issue these companies face is the lack of a minimal source from some populations, those that you mentioned included. Now I agree that it would be logical if all of them used the same clusters or models. Let's see what will happen, as FTDNA has probably been receiving lots of complaints after this update.

JohnHowellsTyrfro
04-13-2017, 06:19 PM
DNALand still hasn't fixed that Ashkenazi issue. It is also a lesser problem with 23&me given the way that their algorithm is arranged. Both mistake certain Southern European genes with Ashkenazi. It likely happens with all the ethnicity genetic tests to some degree hence the ever so common "I didn't know I had Ashkenazi" on various forums.

There was a case study a while back about that and a similar ancestry algorithm as 23&me where a guy was "20% Ashkenazi". Shame when compared against legitimate Ashkenazi he (and likely others) was about as legit as a banana being passed off as a blueberry.

I've scored Ashkenazi a few times and I'm British with no known Jewish Ancestry. Chromo2 stated specifically Ashkenazi and Finnish could just be some similarities in ancestry with other population groups not necessarily A or F. John

Calas
04-13-2017, 10:31 PM
I've scored Ashkenazi a few times and I'm British with no known Jewish Ancestry. Chromo2 stated specifically Ashkenazi and Finnish could just be some similarities in ancestry with other population groups not necessarily A or F. John

It is ancestral. About 5% of Welsh I believe share similar ancestral genes with Ashkenazi if I remember the summary correctly. Same reason why some Scots likely happen to show quite a bit of Finnish when compared to other British populations with no probable recent (non-Viking or older) Finnish ancestry.

curiousII
04-16-2017, 01:43 PM
It is ancestral. About 5% of Welsh I believe share similar ancestral genes with Ashkenazi if I remember the summary correctly.

Hi, Calas! I was following another discussion you were involved in over in the Living DNA German Project thread, spefically the interchange between you and MatAust21. There were a few posts, these towards the end, posts 42-44: http://www.anthrogenica.com/showthread.php?10223-German-DNA-Project-by-Living-DNA&p=227061#post227061

It appears that MatAust21 is Brazilian and fairly opinionated on whether Ashkenazi should be included in the project. I think the points that you debated with him are relevant to what DNA.Land considers when assigning its results to its tests. And FTDNA now, also, with its new myOrigins 2.0. For example, when I click on my Ashkenazi on my map I get an overlay of my Eastern Europe matches. That Ashkenazi is one of two Trace Regions, the other Western Middle East which appears to be the eastern shore of the Med.

Now, going by the arguments posed by those who don't believe Jewish to be included in indigenous German (or European, for that matter) why would my Ashkenazi hit show in Eastern Europe rather than being included in my Western Middle East match? I believe that geographic area is what was the Fertile Crescent and the Levant, isn't it? And that's historically accepted as being the area where the Jews began? Or, if the Jews have been assimilated into the European stock after commingling for so may thousands of years now, is my Western Mid East match superfluous? Which geographic splat better represents that Trace Region: baby blue in the East Med shore, or the dark blue/purple splat over East Europe?

The past few years since I've taken my first autosomal test at a lesser, inferior, trauma-inducing DNA company that refused to assign percentages to the customer's results ("You have heritage from Area X and Area Y, there's no way to assign a percentage to either no matter what your skin or eye color is because you're both. There.") I've read what seems to be an on-going debate about both Ashkenazi and Sephardic heritage. I read that the Sephardic Jews are from the original 12 tribes, the Ashkenazim are now Khazars which are not true Jews. If the latter is true, and my trace hit a real match, would that explain my Asian matches rather than Jew? Aren't the Khazars said to be from an Asian region? Wikipedia, though I'm sure you have much better reference material than I: https://en.wikipedia.org/wiki/Khazars

Longer post than I intended, I'll sum up. Does DNA.Land lump too many regions into the European label? Is "West Euroasian" to vague a label? And how did your discussion on the other thread end?

I have a couple of matches on DNA.Land now, one of whom I also have of my FTDNA FF results. I've emailed him, interesting to see if I get a reply.

edit: You'll see my haplogroups are DF27>Z2573 and H11a, both attributed to European origins. Or, H11a was said to have "sheltered" somewhere in Western Asia/Siberia during the LGM while the rest of the mtDNA haplogroups froze. Or, by other accounts, sprung up in Iberia and spread East after the glaciers retreated. Either, both my y-and-mtDNA haplogroups are European and neither are said to have origins in the Middle East or North Africa. If the trace results in my autosomal tests are good and not statistical noise, then that's heritage my ancestors picked up along the way rather than illustrative of my family's true geographic point of origins. That being said, does it lend ease into determining if a trace region is a real hit or just a false positive? Would data like that affect the study you were talking of on the other forum?

And, if a study like the one you were discussing didn't include both haplogroup and autosomal heritage, how could it even be valid? How could it have any relevant results?