More Recent Comments

Monday, April 11, 2016

My DNA ancestry

I sent a DNA sample off to ancestry.com a few weeks ago and here are the results.


Ancestry has a peculiar way of identifying haplotypes. When they say "Ireland," they mean Ireland and Scotland.1 When they say Great Britain, they mean that they don't distinguish between England, most of Scotland, and most of Normandy.

The results look fairly accurate. My maternal grandmother is Irish—both her parents immigrated to Canada from Ireland in the late 1800s. They descend mostly from English settlers who moved to Ireland in the 1600s. That's why I'm not 25% Irish.

My maternal grandfather is a mixture of English, Scottish, French and Dutch ancestors. The dominant DNA markers should be Scottish and English.

My paternal grandfather is from Russia, near the Volga river north of Volgograd and south of Saratov. He was German ... descended from German immigrants brought in my Catherine the Great in the late 1700s. The German communities on the Volga did not mix genes with the local Russians so the haplotypes will be German. This is mostly why I'm almost 25% German.

My paternal grandmother is from Volhynia in northern Ukraine. Many of her ancestors were German having recently (1700s) settled in the regions from Germany and German-speaking parts of Poland. Those families mixed with the local populations of Poland, Lithuania (now Belarus), and northern Ukraine.

The trace Scandinavian haplotypes could come from either side of my family through Poland/Germany or through Scotland/Ireland.

Ancestry.com identifies all my DNA relatives who have had their DNA tested. There are 73 of them in all but none closer than 4th cousin. Fortunately, some of them seem to related to my paternal grandfather. That's the link I wanted to explore so I'll be getting in touch with them.

If you add in the other bits then I've got all of Europe covered except Italy. That doesn't explain why I like spaghetti & meatballs and pizza.



1. The DNA analysis is done in a lab in Ireland!

44 comments :

Faizal Ali said...

I'm kind of interested in having this done, but not enough to spend $149.

apalazzo said...

Moran = of Moorish decent, non? So where are the Semitic haplotypes? ;)

Bryan said...

Interesting...I wonder if we share a common relative. My paternal grandfather too is a German who was driven to Russia during the Napoleonic wars.

I've heard from others, including family members who are quite interested in genealogy, that anscestry's service wasn't totally accurate - interesting to see a counter-example.

Larry Moran said...

I'm actually quite relieved that I don't have any Italian ancestors. They're not the best people to be related to.

SRM said...

I'm actually quite relieved that I don't have any Italian ancestors. They're not the best people to be related to.

I predict a very long thread emanating from that comment. ;-)

William Spearshake said...

I have an Italian sister-in-law. She is great. Except for the extra five inches on my waistline.

Joe Felsenstein said...

They have had some artistic achievements over the years. Like the Renaissance, for example. Plus there was some guy who went around dropping balls off the Leaning Tower of Pisa, Galileo or something like that. No time here to talk about food. In more recent times, Frank Sinatra, Anthony Fauci, Luca Cavalli-Sforza, not to mention Frederico Fellini.

Joe Felsenstein said...

Your widely-dispersed ancestry is a much more accurate result than one gets from some companies. There are companies that sequence your Y chromosome and purport to tell you exactly which locality you "come from". Aside from the issue of statistical uncertainty, it only gives you a picture of your father's father's father's line. Your very many other ancestors almost certainly could not come from that exact place, unless your more recent ancestors restricted themselves unreasonably in their choice of partners.

Dazz said...

I'm guessing that was some sort of inside joke or something. Or at least I hope so because it would be a huge disappointment to learn that Prof. Moran is a xenophobic moron

Larry Moran said...

I'm responding to a comment from my friend and colleague Alex Palazzo. Can you guess where his family comes form?

Dazz said...

Thanks for clarifying, I suspected it was an inside joke. Sorry about that

Robert Byers said...

The people who settled Ireland in the 1600's were mostly Scottish. A few English but not many. The few English could not but helped breeding with the Scottish.

I question that much can be got from these DNA things. Except a very general celtic or Germanic gene expressions. THEY DO try to say they can narrow it down to Danes, swedes , and so on. HMMM. It seems too precise for such genetic markers.
There was so much inter breeding EVEN before the Romans took over and after that I question the accuracy in these things except basic conclusions.
I also wonder , as a creationist, that options for quick gene change are not possible. So like environment creates like genes. Yet no meaning biological relationship otherwise.
Not sure.

Ed said...

Robert says:
"It seems too precise for such genetic markers."

Does it? And how do you think they can send a criminal to jail based on a DNA sample found at the crime scene? Or do paternity testing?

Do you actually have a clue why the company doing the ancestry testing mention this:
When they say Great Britain, they mean that they don't distinguish between England, most of Scotland, and most of Normandy.

I do seriously doubt it, but perhaps you'd like to give it a try Robert?

Larry Moran said...

I guess it's just an amazing coincidence that the DNA haplotypes in my genome just happen to correspond with the haplotypes common in the area where my immediate ancestors came from. As a creationist, I'm sure you have a good explanation for this. Would you like to share it?

Aceofspades said...

Very cool... so no slaves in your ancestry then?

Like you, I'm mostly Western European but I was surprised to find I had 1.4% Subsaharan African DNA over 3 segments (136MB, 28MB, 154MB) and 0.4% East Asian DNA over 2 smaller segments (12MB, 34MB)

I still haven't found out who these ancestors are.

I'd like to try an predict how many generations back I should expect these ancestors given the number and size of the segments. 23andme will predict that I am 3rd - 4th cousins with somebody if I share 0.6% of DNA with them over 3 segments - but that would be comparing me to somebody who is potentially part of the same generation as me as opposed to comparing me to an ancestor.

I'd appreciate any help here if anyone knows how to do the math.

John Harshman said...

Very cool... so no slaves in your ancestry then?

Oh, probably plenty of slaves, just not any sub-Saharan African ones, and not lately. Slavery is hardly a purely recent phenomenon.

Larry Moran said...

Robert Byers says,

The people who settled Ireland in the 1600's were mostly Scottish. A few English but not many. The few English could not but helped breeding with the Scottish.

As usual, your understanding of history leaves a lot to be desired. Yes, it's true that the majority of new immigrants to Ireland came from Scotland but there were a substantial minority from England, especially in the south.

My English ancestors settled in Tyrone and Fermanagh (now Northern Ireland). They all had English surnames (Job, Foster, Emerson, Moore). They only spoke English and they attended Protestant churches.

Robert, I really don't understand where you're coming from. What was the point of challenging my statement that my Irish ancestors were originally English? Is it related to your false claim that Canada was mostly settled by Yankees from New England?

ElShamah777 said...

I'm actually quite relieved that I don't have any english ancestors. They're not the best people to be related to.

SRM said...

You are unfashionably late to the party 777.

Robert Byers said...

I stabd corrected. I understood almost all were Scots. Then the small number of english, Welsh, Dutch etc would of been assimulated by marriage because the whole group over the centuries would of taken on a Protestant Irish identity. no longer just breeding in their original identities. possibly because people moved around so little the english identity would endure.
I did think you might of not known the settlers were mostly Scots because its common to say it was a English settlement.

same as in cAnada. americans think Ontario was settled by english yet it was settled by a Yankee northern migration that also settled the states this side of the Mississippi.
Then after the 1830's(possibly due to the revolts here) it was a protestant Irish, scottish, migration from Great britain that settled ontario. very few English until the late 1800's and that around the lake in towns/cities.
I never said yanks settled ontario but rather they were 80% of those who settled from the revolution until the 1830's. then heavy migration from Britain.

Robert Byers said...

I understand this and watch it on youtube.
its not about close modern relationships.
i'm questioning the more past ones based on these things.
They admit its all about Celts and Germans in these areas. THEN they say they can track it further. Right down to Danes, Swedes, etc
I question the deeper conclusions.
this because I find the whole thing of DNA tracking to be speculative. Even though true at recent relationships.
Further i question that dNMA is a good track because i see it can be influenced by like conditions to unlike population groups.
This is a big subject.
Yet a example is that all white europeans were already segregated in tribes, speaking different languages, before coming to Europe and getting white.
So whiteness was not from biological breeding relationship but just a envirormental trigger.
Like with White, rabbits, wolves, bears up north.
So the WHITE skin genes would be in everyone but not becaise of bio breeding.
I am sure marsupial wolves are just placental wolves despite having genes like the other "marsupials". SO its easy to see the adaptive triggers bring a like genetic "score" but not from a common descent concept.

They do cool stuff with gebne tracking. They say they can find the Vikings in the North English gene pool. Hmmm. Maybe but they are all the same.
I just question speculate the details in the gene ancestry stuff.

Duane Waite said...

I'm thinking of doing this myself to see if they can find any Neanderthal and/or Denisovan DNA in my genome. Is this something that Ancestry is able to do?

Rolf Aalberg said...

Robert Byers said:I just question speculate the details in the gene ancestry stuff.
With little brains, speculation is all you can do. Scientists are stupid, you know better, science is useless, science is for morons, isn't it? How come they can trace ancestry simply by checking DNA samples? Is it just an irrelevant coincidence that you are stuffed with the DNA of your parents - and not by the DNA of people not in a close relationship with you?

Aceofspades said...

*recent ancestry :P

Aceofspades said...

23andme offers this service. Alternatively, you could take your results from ancestry.com and use this tool:

http://www.y-str.org/2014/12/ancient-calculator.html

Warning: I get 2 very different results from these two different tools so we can't be sure what parameters 23andme use to come up with their results.

Duane Waite said...

Thanks!

Capt Stormfield said...

My wife was born in Ireland and her 23&me results showed a bit of Middle-Eastern. We got the results coincident with media reports last year about the earliest Irish settlers being from that area. Presumably some of her Irish roots go very deep.

I was also pleased to note that she has a bit more Neanderthal than I do.

SRM said...

Aceofspades: any ideas which is the most accurate assessment and why two different methods would yield different results? How different were your results? Kind of intrigued with trying this after this post, but on other hand there is no value in inaccurate information.

Larry Moran said...

You have eight great-grandparents. Each one contributed, on average, 12.5% of your genome. The contribution from each of your great-grandparents makes up 6.25% of your genome. That's about as far back as you can go and still have a reliable estimate of where your ancestors came from.

The contribution of Neanderthal DNA to European genomes is almost fixed. There isn't much variation segregating in this population as far as I know. Everyone has the same Neanderthal markers. It's unlikely that two people with mostly European ancestors will differ in the amount of Neanderthal DNA. I would be very skeptical of any company that reported a difference

Joe Felsenstein said...

The contribution of Neanderthal DNA to European genomes is almost fixed. There isn't much variation segregating in this population as far as I know.

I'm not so sure. My colleagues Josh Akey and Ben Vernot conclude otherwise.

John Harshman said...

Hey, if, as hypothesized, some of that neandertal DNA is under selection, shouldn't we expect a bit of a north-south frequency cline?

Joe Felsenstein said...

I don't think this has been looked at yet, except of course for the fact that Neanderthal admixture is not much into Africans. We do expect to see "selective sweeps" of Neanderthal or non-Neanderthal sequences through regions of the human sequence, where there is selection favoring one or the other.

Anonymous said...

This is why I like having Robert Byers comments. I learn things I'd never have thought anyone thought.

Unknown said...

I have to say I'm disappointed that he mentions polar bears, yet fails to deliver his classic catchphrase.

SRM said...

I have to say I'm disappointed that he mentions polar bears, yet fails to deliver his classic catchphrase.

Its just a line of reasoning?

Well, I can attest to his reasoning as I have observed that I am much whiter in the winter and more browner in the summer. Who can tell what I would look like if I spent winters in Florida. Surely that would confound the ancestory services.

Unknown said...

"Polar bears are scared to the point of white hair."
It's to Byers what the Banana is the Ray Comfort. Or fishing hooks to Adnan Oktar.

AllanMiller said...

I think he's onto something. I do find that all the animals in a region are exactly the same, apart from the differences. It's like dogs that look like their owners. Only DNA.

Aceofspades said...

I have some idea why the two results are different and that has to do with the parameters used.

23andme tell me that I am 3% Neanderthal which is above average for their European members. How they come up with that number though is completely opaque.

When I run the app I linked to however I am able to tweak the parameters.

Here is the result of my comparison with Vi33.26

http://i.imgur.com/7SAQUkX.png

If I increase the allowed errors per segment to 2 my percentage match rises to 2.86%

http://i.imgur.com/fiV99Zh.png

If I decrease the "total SNPs per DNA segment" from 60 to 50, my amount of shared DNA rises to 2.17%

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

Unknown said...

I don't know how opaque 23andme's analysis is. The estimator program was created by Eric Durand who has a good background in genome analysis. His white paper on the genome estimator can be found at 23and me site as well as online, e.g. http://stanford.edu/class/gene210/files/readings/23andme_Neanderthal_Ancestry.pdf

I trust its results infinitely more than I would trust the y-str.org results which is run by a fundie xian computer programmer.

Unknown said...

I should have also referenced a little known online neandertal calculator from Interpretome at http://www.interpretome.com/

They only accept 23andme files (and 1 other company I don't know).
Benefit is that they list the 42 SNPs (84 alleles) they use for calculation, so I would guess any one who has their SNP data and this list of SNPs might be able to make a rough calculation. For example, according to 23andme calc. I am 3.1% which is 97th-98th percentile for Europeans. At Interpretome this shows me having the neandertal allele in 15 of 84 positions.
If Larry approves I will post the 42 SNP table... I suspect they can be found online, e.g at ISOGG.


The "grain of salt" admonition applies here: Most of these tests are "tentative" and just for fun. They might improve as we get more neadertal genomes sequenced.

Unknown said...

Here are the SNPs and relevant alleles listed in SNPedia: https://www.snpedia.com/index.php/Neanderthal
Took me some time to find, but with a little bit of work, anyone with their genome data should be able to do some calculations. Sorry for the several posts.

Aceofspades said...

> which is run by a fundie xian computer programmer

I didn't know that. I'm not sure why it makes a difference though so long as he understands the concepts. The point I'm making though is that it seems to be the case that there are different ways of coming up with different numbers depending on how you set parameters.

I had tried interpretome before and it told me I matched 18 out of 84 positions. If this could be extrapolated to the rest of my genome it would make me 21% Neanderthal which would be ridiculous. 23andme tell me I am 3% Neanderthal. My concern with interpretome is that they only use 42 SNPs to draw their conclusion.

W. Benson said...

Most certainly have ancestries with slaves, slavers, cannibals, and horse thieves.

Unknown said...

Ace - except you don't simply extrapolate across your whole genome: you base it only on the 42 SNPs that have been concluded to be extant in neandertals at the time these calculators were created. Since SNPedia still lists (updated Jan 2016) these 42, I assume that base number has not changed.

Here is how you would do the extrapolation if we accept that the range of neandertal in modern Europeans is 1 - 4 %. (see http://ww2.kqed.org/quest/2011/08/29/how-neanderthal-are-you/ ) If you have a low number of these alleles (say 5-7 of 84) you should tend toward the 1%, and if you have a high number (say 20-25) you should tend toward the 4%... NOT 21%. Interpretome no where claims that they are giving you a percentage.