Ancestry Autosomal Results

[auntsha]

Results

My husbands results:
47%. Eastern Europe
22%. British Isles
13%. Southern Europe
11%. Scandinavia
7%. Persian/Turk
Caucasus
He overwhelmingly has gotten and is getting" European Jewish" matches at Ancestry. Is there a survey to designate what they were And that is why they are so specific? His one great grandmother was from Austria and we think she is the one giving all these matches. There is such a large % of East Eur??
He is German and Swiss is that under the British Isles or Scandinavia or is it counted in the East European? He is not British or Irish so heavens knows where that comes from. He has maybe 1/8 or so Italian so the S Eur is ok and the ME McDonald thought from FTdna raw data was Jewish. It was not mentioned to him at all. I feel they got everything but just where to put each ethnicity under is a problem.

[Loreli]

So far, I've not found any true reason to quibble about the Ancestry results. While I don't have a tree to bounce of off, the ethnicity is not crazy out of whack from what I got here, afterwards. I'm expecting quite a bit of German Ancestry once I do figure out my tree, along with whatever else is in there. (And for those of us without trees, Ancestry is at least a good way to check your "matches" trees so you can hazard guesses.)

Ancestry breakdown:

Central European
84%
Scandinavian
11%
Uncertain
5%

FTDna breakdown:

Continent (Subcontinent)
Europe (Western European)

Population Percentage
98.98% (French, Orcadian, Spanish)

Margin of Error
±4.38%

[ragnar]

I am not real happy with my Ancestry ethnicity breakdown. 43% British,41% Central European,9% Eastern European,7% unknown. I gave them feedback that I have 7 of my great great grandparents being Norwegian, traced back to the 1500's & 1600's in Norway. No response. I realize that some Scandinavian has been included into British, however my fear is that some of Norwegian (partial or full descent) matches may skip over me since I don't show any percentage for Scandinavia. I have found that 3 of the very distant Ancestry dna cousins are related on one of those Norwegian lines. The only reason I was able to find the connections was by checking person by person in each of my dna cousin's trees for Norwegian people that came from my ancestor's Norwegian locations. I happen to have bygdeboks (compiled farm histories for select parishes in an area) for 3 of the areas & was able to find their ancestors in the books & then find the relationship to me. I was able to extend those particular lines for those dna cousins (sent the info to them). Ancestry has so far given me 2 hints to 2 Norwegian dna cousins (even though no Scandinavian % for me!). They do not know what they are doing regarding Scandinavian names. They kind of get the bit about 1st names - but they don't quite know what to do with them. They also don't understand that the location is one of the primary things that needs to be considered, in relation to naming patterns. For an extreme example, the latest "hint", was that my Hans so & so (in Norway) may be the same as my dna cousin's Hans so & so (in Sweden), (& also ignoring their Norwegian ancestors, which is where I found the actual match). Ha! Unbelievable!

[djknox]

Ragnar - admixture results, whether at Ancestry or FTDNA, amount to little more than pulling data from one's arse. The sophistication just isn't there yet. Maybe after Gen 2.0 has completed and trickled down will there be some more meaningful output. Take a look at the thread under family finder: "How are reference populations chosen" - my 2c worth are on posts 29 & 31. Of course in defence of admixture tools, its not an easy thing to sort out meaningful reference populations.

Northern Europeans in particular cannot be sorted into useful sub groups. I believe MDexter tried to explain this some months ago. The question is whether they will EVER be able to discern a Viking from a Pict from a Celt? No one hazards to take a guess at that.... certainly not all the regular blogs out there. Just keep spending time & money and hope for the best!

As for Ancestry HINTS.... they're very non-discerning. Ancestry.com is a toy for many... and ethnicity output sheer entertainment. FTDNA takes things more seriously... but whether that leads to much real genealogy progress for most is debatable.

[ragnar]

Thanks, djknox. I will try to relax now. Soothing sounds...mmmmmmmm

[mixedkid]

I am not real happy with my Ancestry ethnicity breakdown. 43% British,41% Central European,9% Eastern European,7% unknown. I gave them feedback that I have 7 of my great great grandparents being Norwegian, traced back to the 1500's & 1600's in Norway. No response. I realize that some Scandinavian has been included into British, however my fear is that some of Norwegian (partial or full descent) matches may skip over me since I don't show any percentage for Scandinavia. I have found that 3 of the very distant Ancestry dna cousins are related on one of those Norwegian lines. The only reason I was able to find the connections was by checking person by person in each of my dna cousin's trees for Norwegian people that came from my ancestor's Norwegian locations. I happen to have bygdeboks (compiled farm histories for select parishes in an area) for 3 of the areas & was able to find their ancestors in the books & then find the relationship to me. I was able to extend those particular lines for those dna cousins (sent the info to them). Ancestry has so far given me 2 hints to 2 Norwegian dna cousins (even though no Scandinavian % for me!). They do not know what they are doing regarding Scandinavian names. They kind of get the bit about 1st names - but they don't quite know what to do with them. They also don't understand that the location is one of the primary things that needs to be considered, in relation to naming patterns. For an extreme example, the latest "hint", was that my Hans so & so (in Norway) may be the same as my dna cousin's Hans so & so (in Sweden), (& also ignoring their Norwegian ancestors, which is where I found the actual match). Ha! Unbelievable!

I understand your concern about the ethnic results from DNA tests. My father is part-Norwegian and my mother, part-Swedisih. Population Finder through FTDNA found none of that (for me or my parents!) but our distant cousin matches tell a different story. Nordic cousins for my father are mostly Norwegian citizens and for my mother, Swedish and Finnish citizens. I am hopeful that my impending results from Ancestry will at least bring up "Scandinavian" with some "Finnish."

Some things I learned from my Nordic Family cousins still in Europe too: Modern populations of Norway or Sweden are probably a combination of the minority Saami and Finnish people and the more dominant germanic Norwegian and Swedish populations. "Forest Finns" and other such Finnish settlers especially must be taken into account when considering modern Norwegian genetics -- especially for those Norwegian citizens who were our immigrant ancestors to North America. More modern Danes and Swedes once controlled Norway, so their genetic input must also be considered when considering Norwegian genetics. And European Norway was once larger than it is today. Borders shifted and some Swedish citizens may actually be born in areas that were once part of Norway. "Shifting borders" can also be an issue that comes up with Danish and German genealogy and genetics.

[tomcat]

Ironic that others without marked Scandinavian ancestry have complained of Scandinavian scores whereas as those with majority Scandinavian ancestry get a poor read on their ethnicity.

Beta get better.

[T E Peterman]

T Peterson is actually Timothy Peterman.

Any chance these Sparks families connect to a Solomon Sparks of Wilkes Co., NC? He was my g-g-g-g-g grandfather.

Timothy Peterman

[MMaddi]

Ragnar - admixture results, whether at Ancestry or FTDNA, amount to little more than pulling data from one's arse. The sophistication just isn't there yet. Maybe after Gen 2.0 has completed and trickled down will there be some more meaningful output. Take a look at the thread under family finder: "How are reference populations chosen" - my 2c worth are on posts 29 & 31. Of course in defence of admixture tools, its not an easy thing to sort out meaningful reference populations.

For another view on the subject of admixture results, sort of an answer to djknox's post 31 in the thread he referenced (http://forums.familytreedna.com/show...5&postcount=31), read my post in that thread (#44 at http://forums.familytreedna.com/show...0&postcount=44). Specifically, djknox writes "However, as recombination dilutes dna quite quickly... ie after several generations traceability is lost due to segments being too 'fractured', it is not clear to me at all when one says that my admixture is 90% British... what that really means in terms of time frames and what 'being British' really means. I believe Matt once explained that the more a population is intermixed in recent generations, the less likely a distinct population reference can be assigned. I think this is the problem with 'British'.. everyone's dna seems very similar due to broad regional intermixing."

My reply in post 44 links to the FTDNA FAQ for Population Finder, which states clearly that PF can assign admixture percentages representing up to 2,000 years of admixture components. djknox refers to segments being lost through recombination over generations. That's true when you are trying to find cousins in a database, since that sort of matching is based on segments of a minimum length. However, in admixture analysis, the algorithm is not looking for segments, but for discrete SNPs, ancestry informative markers (AIMs), which occur at different percentages among different ethnic and geographic populations.

It's true, as djknox quotes mkdexter, that the more admixed someone is, the more difficult admixture analysis is, which I also state in my post. djknox uses the example of "British" and how many different ethnicities come into play in the British Isles, going back 2000 years. (Yet the People of the British Isles scientific project - http://www.peopleofthebritishisles.org/ - has published maps showing very discernible areas of the UK which differ autosomally. So, even "British" can be broken down using living populations.) But for those who have only a couple of major ethnic/geographic populations in their genome, admixture analysis can be quite accurate. So, it's not as bleak as djknox portrays it - as he so colorfully puts it, it doesn't all "amount to little more than pulling data from one's arse." (Tell this to the population geneticists who regularly publish studies on breakthroughsin recent years in admixture analysis. I'm sure they would be amused by his colorful analysis.) That's just his way of discouraging anyone from taking the subject seriously.

[MMaddi]

It's true, as djknox quotes mkdexter, that the more admixed someone is, the more difficult admixture analysis is, which I also state in my post. djknox uses the example of "British" and how many different ethnicities come into play in the British Isles, going back 2000 years. (Yet the People of the British Isles scientific project - http://www.peopleofthebritishisles.org/ - has published maps showing very discernible areas of the UK which differ autosomally. So, even "British" can be broken down using living populations.)

Someone has posted an image of the map I referenced above showing the discernible genetic geographic clusters in the UK - http://pinterest.com/pin/32721534764215309/

If biogeographic ancestry is discernible at the level of regions within the UK, then there is hope that future developments will improve admixture analysis on a continental scale.

[djknox]

Hope for admixture testing

Agreed... there IS hope.... and I don't claim to know anywhere near enough to be able to accurately predict what the future may hold on this topic. I find it curious however that those expert population geneticists you reference so admirably are also quite mute on what the great potential (potential = hope in my world) actually is? Just MAYBE THEY would like to weigh in on this matter? Where are THEIR Posts?

Regardless, MMaddi, nothing you stated above refutes my "colorful observation" that admixture results being produced TODAY, especially for those of N European and/or BRITISH ancestry (as this is where I choose to personally focus), amount to not much more than entertainment value only.

Maybe you could tell me what "DISCRETE SNPs" are aligned to being of Pict, Celtic, Saxon etc. ancestry? Telling someone who is British that there admixture is British isn't worth 100s of dollars to a consumer. No wonder Ancestry, 23&Me and FTDNA databases are full of British clientele ! As for autosomal testing in general, i've got ZERO hits to someone in the UK. So familyfinder, dna.ancestry etc. are not much good for matching anything beyond some quaint American colonial heritage... whoopy!

I sincerely invite a population geneticist to join these forums and provide real insight to why people should be spending money on these tests. Spencer Wells geno project is commendable, and equally ingenious by getting folks to self finance the project. I don't belittle the scientific complexity and efforts involved, only the commercial admixture products on offer - those that claim to tell one what their geographical ethnicity is. If consumer protection agencies had even a remote abillity to understand anything of subtlety (which they don't), they would be all over admixture products. Alas such products are way above their punching weight. This is why I advocate that it behooves commercial providers to give some insight to what value one may extract from the tests, especially wrt forthcoming advancements around the corner. I am not a geneticist (obviously) but I know enough about scientific research and commercial applicability to know that what is to be offered 5 years from now is generally well in progress today. So what will be on offer in the next 2 years is DEFINITELY known today... yet all these forums, and the popular genetic ancestry blogs, conveniently don't even breathe a peep about what's around the corner.

As you are such a strong advocate of admixture analysis, maybe you could share with everyone why that is? Without any intention to disparage, either there really isn't anything around the corner worth talking about, or there's a bad strain of NKS syndrome in the blogosphere. I've seen a few fleeting mentions about new and improved FTDNA tools, yet no press releases or announcements to the public of immanent product enhancements or, less I be too optimistic, new products altogether.

By the way, whatever became of the ftdna newsletter? The latest available on this website is 18 months old. For a field supposedly developing so so fast, that's quite an interim of silence. In the real world, companies selling to consumers create excitement about their ongoing product upgrades.

[Mlawton]

Thank you all for helping me understand the probability statistics better. It is making more sense to me. I still think she should have some distant Sparks matches even if not Traylor, but.....

Yaffa, thank you and yes, these are my Sparks you mention, NC, GA, MS, AR TN...those are the states I have traced my line so far, but yes, they definitely went forth and multiplied!

Timothy, which Solomon are you descended from? There are a quite a few of them. Would it be the son of Joseph from Frederick MD? If so, yes, we are related through William Sample Sparks, who was a first cousin (I think).

[djknox]

Thanks, djknox. I will try to relax now. Soothing sounds...mmmmmmmm

EXACTLY! Except for the small matter that i'm burning lots of money chasing a ghost. Where's Bill Murray and Dan Ackroyd when one needs them lol!

I do hold out with the hope that admixture and cousin matching will all one day start to come together... as knowing one's regional ancestry beyond the lmited paper trail would be so so cool! On this at least likely everyone agrees.

[djknox]

Thank you all for helping me understand the probability statistics better. It is making more sense to me. I still think she should have some distant Sparks matches even if not Traylor, but.....

Yaffa, thank you and yes, these are my Sparks you mention, NC, GA, MS, AR TN...those are the states I have traced my line so far, but yes, they definitely went forth and multiplied!

Timothy, which Solomon are you descended from? There are a quite a few of them. Would it be the son of Joseph from Frederick MD? If so, yes, we are related through William Sample Sparks, who was a first cousin (I think).

Without analyzing your tree and fully understanding your specific case, I think you are the best to figure out if something is amiss or not... just bear in mind that after 1/2 dozen generations any one individual could have lost much of a particular line in their autosomal dna (at least this is what we're told)... so it is perfectly understandable if one of your Sparks 4th cousins just doesn't match with the others... good luck.

[Yaffa]

Thank you all for helping me understand the probability statistics better. It is making more sense to me. I still think she should have some distant Sparks matches even if not Traylor, but.....

Yaffa, thank you and yes, these are my Sparks you mention, NC, GA, MS, AR TN...those are the states I have traced my line so far, but yes, they definitely went forth and multiplied!

Timothy, which Solomon are you descended from? There are a quite a few of them. Would it be the son of Joseph from Frederick MD? If so, yes, we are related through William Sample Sparks, who was a first cousin (I think).

There is also a possibility your 4th cousin paper is wrong, there could also be in "Oops" in their line or you just inherited different DNA like I explained in my previous post. Would really help if and when ancestry releases raw data. Adding X Chrome matching would also help because this is strictly mother lines. As your alleged cousin gets more matches they will know if they are matching others in their family that they are supposed to match on paper or not. Your cousins best bet is to test known closer family in their lines.

My ancestors come from the area of Timothy's. The Sparks family around my family may be his though I dont have any Sparks matches at FF. Ancestry test is pending right now. The Sparks family took over our area like the Combs family and they are probably not all related even though people think they are. I think there are just as many Sparks as there are Combs. Anderson is another one and they all really need to DNA test to sort them out. Just for example our area I know there are 4 different Cook lines that are not related, several Martin lines that are not related etc.