Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
No, and I am not saying some of the samples couldn't be from people who live in other places but with 100% Italian ancestry. My point is, there are very probably also samples from Italy, since the reference panel has nothing to do with where AncestryDNA sell their tests. Additionally, your assumptions about the size of the population in the reference panel are incorrect.
The reference panel population is fairly small. As I stated. You need about 250 samples for a single marker so 1000 isn't all that useful.
Also unfortunately there is no way of knowing if someone has "100% Italian" ancestry. That's kinda the point. They're trying to find reliable proxies...in this case that is people with some acceptably sufficient documented ancestry in a specific Geo. That's not the same thing as sampling that population directly. Its a proxy.
The reference panel page doesn't have enough information about its methodology to identify whether or not it has samples collected directly from Italy; my point is you're assuming it must...which there is no reason to do.
According to the page, their reference panel is ~2/3 derived from clients of Ancestry DNA. That means, that contrary to what you state, it has a LOT to do with where they do business. It's certainly possible that there are sample collected from Italy in the other cohorts...but they don't go into the details of those.
Yes, I think despite AncestryDNA's attempts to narrow down the populations, I feel that in many cases, we still need to view the results on a broader level. My mom has a lot of German/Swiss ancestry, but in the update she gets no results in Germanic Europe at all. I could throw a hissy fit, rant about how unreliable the update is and what a joke that makes AncestryDNA, or I could look at the rest of her results and see that she gets very high amounts in two neighboring regions - one in particular she gets much higher results in than should reasonably be expected for her known ancestry. So it seems her Germanic ancestry is just showing up in a neighboring region, which was common for many people before the update too, and at other companies, and will continue to be common for the foreseeable future. This is not exclusive to AncestryDNA. If you feel that this makes the ethnicity estimate (in general, from any company) a waste, then fair enough - I've always maintained that the ethnicity estimate is fun to explore, but the true value of the test is with the DNA matches.
Quote:
Originally Posted by Einhander
That's what I feel.
As do I. We've discovered two children of my oldest brother no one knew about. We think he might have known about one, but not the other.
On Ancestry's Facebook page I'm seeing many people with Spanish ancestry saying they now show zero percent Iberian and instead are Irish and British.
Apparently the Basque people are related to the Irish people.
So this is not weird. I only assume that many centuries ago, irish fishermen settled in the atlantic coast of northern Spain/
As I mention many times: Europe is just like a big Nation and people have been moving from one corner to another for centuries.
Apparently the Basque people are related to the Irish people.
So this is not weird. I only assume that many centuries ago, irish fishermen settled in the atlantic coast of northern Spain/
As I mention many times: Europe is just like a big Nation and people have been moving from one corner to another for centuries.
I thought the Basque were a very distinct DNA ethnic group. So how would they come up as Irish?
Apparently the Basque people are related to the Irish people.
So this is not weird. I only assume that many centuries ago, irish fishermen settled in the atlantic coast of northern Spain/
As I mention many times: Europe is just like a big Nation and people have been moving from one corner to another for centuries.
Research Black Irish.
People of Spanish descent that ended up in Ireland for a number of reasons. More than once iirc.
My wife and I both lost the Iberian Peninsula link from ours.
i dont care for it because I was 1% this and 7% that, and 10% of this, Now Im 100% british, and that I dont agree with because the british came from somewhere
The reference panel population is fairly small. As I stated. You need about 250 samples for a single marker so 1000 isn't all that useful.
Lol, then none of the reference panel is useful. Fair enough if that is your argument, but you were clearly originally trying to claim that Italy in particular had a smaller population in comparison to the rest of the reference panel, which is not the case.
Quote:
Also unfortunately there is no way of knowing if someone has "100% Italian" ancestry. That's kinda the point. They're trying to find reliable proxies...in this case that is people with some acceptably sufficient documented ancestry in a specific Geo. That's not the same thing as sampling that population directly. Its a proxy.
All samples, regardless of where they are from, require the same level of acceptable documented ancestry, so it really doesn't matter where they are from. If someone has documented ancestry entirely in Italy going back to the 1700s, what does it matter if they are from the US or Italy? If they're from the US, their ancestry is going to be the same as someone's who is from Italy because the requirements for documentation are the same.
My point is, there is no reason to think that the sample population for Italy is weaker than any others, which was your initial argument.
Quote:
The reference panel page doesn't have enough information about its methodology to identify whether or not it has samples collected directly from Italy; my point is you're assuming it must...which there is no reason to do.
Okay, but you were assuming that it can't/doesn't. Isn't that the same thing?
Quote:
According to the page, their reference panel is ~2/3 derived from clients of Ancestry DNA. That means, that contrary to what you state, it has a LOT to do with where they do business. It's certainly possible that there are sample collected from Italy in the other cohorts...but they don't go into the details of those.
Where did you read that?
I guess you are referring to this:
"In developing the most recent AncestryDNA reference panel, we began with a candidate set of close to
34,000 samples. First, we examined over 1,000 samples from 52 worldwide populations from a public
project called the Human Genome Diversity Project (HGDP) (Cann et al. 2002; Cavalli-Sforza 2005),
together with over 1,800 samples from 20 populations from the 1000 Genomes Project (McVean et al.,
2012). Second, we examined samples from a proprietary AncestryDNA reference collection as well as
AncestryDNA samples from customers. Most of the candidates were selected from the last two groups
only after their family trees confirmed that they had a long family history in a particular region or within a
particular group. A small number of candidates were selected without a deep family tree but these passed
the rigorous vetting process outlined below. Although it was not possible to confirm family trees for
HGDP and 1000 Genomes Project samples, these datasets were explicitly designed to sample a large
set of distinct population groups representing a global picture of human genetic variation."
However, they do not give the numbers for the amount of samples they took from AncestryDNA customers. They say "Second, we examined samples from a proprietary AncestryDNA reference collection as well as AncestryDNA samples from customers." Those proprietary AncestryDNA reference collection do not come from AncestryDNA customers, and we don't know how many of which there were. So no, we do not know that 2/3 came from customers.
But why does that make you lose trust in AncestryDNA? The ethnicity estimate has always been an estimate, and what I said about it being fun to explore but the true value of the test being with DNA matches has always been true. And it applied to every other DNA company out there. So why lose trust in AncestryDNA, and only AncestryDNA, now?
I thought the Basque were a very distinct DNA ethnic group. So how would they come up as Irish?
They have blue eyes, dark hair.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
