World Families Forums - Some Comparisons By Country of Numbers Tested

Welcome, Guest. Please login or register.
September 23, 2014, 07:21:38 AM
Home Help Search Login Register

+  World Families Forums
|-+  General Forums - Note: You must Be Logged In to post. Anyone can browse.
| |-+  R1b General (Moderator: rms2)
| | |-+  Some Comparisons By Country of Numbers Tested
« previous next »
Pages: [1] Go Down Print
Author Topic: Some Comparisons By Country of Numbers Tested  (Read 1333 times)
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« on: December 05, 2009, 03:19:37 PM »

I think we all  (well, most of us, anyway) realize that the North American dominance of commercial genetic testing results in a skewing of results toward the British Isles. Since the British Isles are predominantly R1b1b2, this bias really affects how we view the various R1b1b2 subclades. I want to take a look at how some of this plays out in terms of actual numbers. So, I am going to use the numbers from FTDNA's "Ancestral Origins" pages and population estimates from this web site to try to measure the relative proportions of men who have been tested to at least 12 y-dna markers. That  last web site's figures are based on United Nations population estimates of July 1, 2009.

One slight problem with using FTDNA's Ancestral Origins pages is that those numbers are constantly going up. But we have to try, and, given the huge gap between the British Isles numbers and those of other places, it doesn't seem likely the relative proportions will change much overnight. Another problem is that Ancestral Origins figures include members of the diasporas of the various nationalities named. I looked for a convenient, all-in-one-place source for estimates of the populations of various national diasporas, but I couldn't find one. So, we'll have to make do with the limitation that diasporal populations are not included in UN estimates of national population figures.

Of course, my interest is in R-L21*, so I am looking at these numbers in terms of R-L21*. But the numbers are good for thinking about the proportions of any haplogroup or clade found in both the British Isles and elsewhere in Europe.

Let's first compare the Republic of Ireland (referred to hereafter simply as Ireland) with France, since both places appear to have a high frequency of R-L21*.

Ireland's estimated population is 4,459,300. Guessing that about half of that figure is male gives us 2,229,650. As of right now, FTDNA's Ancestral Origins pages report that 11,384 men of Irish descent have been y-dna tested to at least 12 markers. For France, those figures are 65,073,482 in estimated total population, of which about 32,536,741 would be males. Ancestral Origins says that as of right now 2,764 Frenchmen have been y-dna tested to at least 12 markers.

How does that work out?

Ireland

Approximate Male Population of Ireland =  2,229,650

Irish 12-marker+ Tests = 11,384

Proportion Tested = 11,384/2,229,650 = 0.005105734

France

Approximate Male Population of France = 32,536,741

French 12-marker+ Tests = 2,764

Proportion Tested = 2,764/32,536,741 = 0.00008495

Let's compare those two proportions.

Irish Proportion Tested = 0.005105734

French Proportion Tested = 0.00008495

Note that the French figure is by far the smaller of the two.

In terms of absolute numbers, four times as many men of Irish descent have been tested as men of French descent, but in terms of relative populations, for every 511 Irishmen who have had y-dna tests to at least 12 markers, only 8 Frenchmen have been so tested. That's almost 64-to-1!

So, let me ask you: Do we have 64 times the Irish R-L21* results than we do French R-L21* results?

That was a rhetorical question. The obvious answer is no. We only have just under 6 times the number of Irish results versus French results. That means that French R-L21* results are actually higher than Irish R-L21* results in terms of relative proportions tested of the male populations of those two countries.

I'm not arguing that France will turn out to have a higher frequency of R-L21* than Ireland, although that is possible. Ireland is a relatively small island nation, where founder effect and genetic drift can be more profound than in a larger region like France. What I am trying to show is that the rush to conclude that L21 might have originated in Ireland was, to put it very mildly, premature.




« Last Edit: December 05, 2009, 03:27:26 PM by rms2 » Logged

NealtheRed
Old Hand
****
Offline Offline

Posts: 930


« Reply #1 on: December 05, 2009, 03:26:38 PM »

Thank you for taking the time and providing that analysis, Rich. It provides insight into how biased the sampling frame is in the databases.
Logged

Y-DNA: R-Z255 (L159.2+) - Downing (Irish Sea)


MTDNA: HV4a1 - Centrella (Avellino, Italy)


Ysearch: 4PSCK



rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #2 on: December 05, 2009, 03:48:49 PM »

You're welcome, Neal.

For Scotland the total population estimate (I had to go outside the UN figures for it) is 5,168,500. If about half of that is male, that would be 2,584,250. Ancestral Origins says that as of right now 9,110 Scotsmen have been tested to 12 markers or more.

Here's how that works out.

Scotland

Approximate Male Population of Scotland = 2,584,250

Scottish 12-marker+ Tests = 9,110

Proportion Tested = 9,110/2,584,250 = 0.003525201

That's a slightly smaller proportion than for Ireland, but it's far larger than the French proportion. Relative to their respective populations, for every 353 Scots who have been tested to 12 or more markers, only 8 Frenchmen have been so tested.

Here's what we have thus far in terms of numbers tested relative to national population estimates (per 100,000 males).

12-marker+ Tests (Regardless of Haplogroup) -

Ireland -     511
Scotland -  353
France -         8


We can see that, relative to its population, which is much larger than that of Ireland and Scotland combined, France is drastically under tested.

And, yeah, that does make a difference.
« Last Edit: December 05, 2009, 05:08:16 PM by rms2 » Logged

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #3 on: December 05, 2009, 03:57:04 PM »

Oh, by the way, in my opening post I did not include the 577 men that Ancestral Origins reports for Northern Ireland in my figure for Ireland, even though I make no difference between the two in the categories on the Y-DNA Results page of the R-L21 Plus Project.
Logged

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #4 on: December 05, 2009, 04:32:48 PM »

Here are the figures for Germany as of today.

Germany

Approximate Male Population of Germany =  40,941,171

German 12-marker+ Tests = 9,948

Proportion Tested = 9,948/40,941,171 = 0.000242983


Here again is what we have thus far in terms of numbers tested relative to national population estimates (per 100,000 males).

12-marker+ Tests (Regardless of Haplogroup) -

Ireland -      511
Scotland -   353
Germany -    24
France -          8



« Last Edit: December 05, 2009, 05:09:45 PM by rms2 » Logged

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #5 on: December 05, 2009, 04:44:47 PM »

Here again is what we have thus far in terms of numbers tested relative to national population estimates (per 100,000 males).

12-marker+ Tests (Regardless of Haplogroup) -

Ireland -      511
Scotland -   353
Germany -    24
France -          8


I realize this method is not perfect, since we lack figures for the diasporas of the various nations, but their relative proportions should perhaps not be too far off the proportions of their current estimated populations.

By way of interpretation, what this means is that, if Ireland, Scotland, Germany and France all had exactly the same frequencies of R-L21*, we should expect 511 Irish R-L21* for every 353 Scottish, 24 German, and 8 French R-L21*. (But that is nothing like what we actually have.)

Sort of puts the Irish and Scottish results in a different light, doesn't it?
« Last Edit: December 05, 2009, 05:10:17 PM by rms2 » Logged

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #6 on: December 05, 2009, 04:57:03 PM »

Let's throw in England:

England

Approximate Male Population of England =  25,600,000

English 12-marker+ Tests = 19,703

Proportion Tested = 19,703/25,600,000 = 0.000769648


Here again is what we have thus far in terms of numbers tested relative to national population estimates (per 100,000 males).

12-marker+ Tests (Regardless of Haplogroup) -

Ireland -       511
Scotland -    353
England -       77
Germany -     24
France -           8



« Last Edit: December 05, 2009, 05:10:55 PM by rms2 » Logged

alan trowel hands.
Guru
*****
Offline Offline

Posts: 2012


« Reply #7 on: December 05, 2009, 08:47:30 PM »

There is no doubt that some people have been blinded by numbers and ignored the sampling biase.  The thing is I think this was obvious a long time ago.  I kept my eye only on the hit Rate for L21 testing in France and ever since the first 6 or 7 tests went in a roughly 50% hit rate has been apparent.  I couldnt understand those who just looked at totals and ignored the huge sampling bias.  I can only guess it was down to the genealogically orientated being desperate to find an isles specific marker but it was a mass logic bypass.  In fact, that this was the main interest in L21 for many is shown by the striking lack of interest in L21 on rootsweb etc where it kind of feels like L21 was blanked as it has emerged more and more clearly it was not an isles-specific clade.  Meanwhile the rush to get more downstream localised SNP continues apace.  I think it is just a fact of life that most in the hobby are looking for country-specific certainty and essentially a more shallow time perspective.  Their main aim is to find ever more localised and even semi-private SNPs that indicates their yDNA origin very closely.  Problem is for those interested in deeper time is that localised SNPs do not tell us anything about migrational history.        
« Last Edit: December 05, 2009, 08:48:32 PM by alan trowel hands. » Logged
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #8 on: December 05, 2009, 08:53:25 PM »

There is no doubt that some people have been blinded by numbers and ignored the sampling biase.  The thing is I think this was obvious a long time ago.  I kept my eye only on the hit Rate for L21 testing in France and ever since the first 6 or 7 tests went in a roughly 50% hit rate has been apparent.  I couldnt understand those who just looked at totals and ignored the huge sampling bias.  I can only guess it was down to the genealogically orientated being desperate to find an isles specific marker but it was a mass logic bypass.  In fact, that this was the main interest in L21 for many is shown by the striking lack of interest in L21 on rootsweb etc where it kind of feels like L21 was blanked as it has emerged more and more clearly it was not an isles-specific clade.  Meanwhile the rush to get more downstream localised SNP continues apace.  I think it is just a fact of life that most in the hobby are looking for country-specific certainty and essentially a more shallow time perspective.  Their main aim is to find ever more localised and even semi-private SNPs that indicates their yDNA origin very closely.  Problem is for those interested in deeper time is that localised SNPs do not tell us anything about migrational history.        

That's right. The current focus on downstream SNPs is kind of frustrating to me. Tests for them are being returned with lightning speed while good ol' L21 tests seem to be taking the full time predicted and more.

I got so excited by L21 (and I still am excited by it) that I must admit further downstream SNPs don't have that much appeal (unless one were found that split L21 in two or something like that). I'm still thrilled with the quest to see how L21 is distributed in Europe, and I have a hard time understanding those L21+ guys who don't share that interest.
Logged

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #9 on: December 05, 2009, 08:58:37 PM »

I guess one thing that can be said of new SNPs downstream of L21 is that they might be able to establish something of a trail that could tell us a bit of the migration story of L21.
Logged

GoldenHind
Old Hand
****
Offline Offline

Posts: 731


« Reply #10 on: December 05, 2009, 09:10:39 PM »

Let's throw in England:

England

Approximate Male Population of England =  25,600,000

English 12-marker+ Tests = 19,703

Proportion Tested = 19,703/25,600,000 = 0.000769648


Here again is what we have thus far in terms of numbers tested relative to national population estimates (per 100,000 males).

12-marker+ Tests (Regardless of Haplogroup) -

Ireland -       511
Scotland -    353
England -       77
Germany -     24
France -           8




Just to complicate things, many English list their country of origin on FTDNA as the United Kingdom, which actually includes Wales, Scotland and Northern Ireland. I doubt the Welsh, Scots and Irish are as inclined to do so, but I could be wrong. Is Great Britain also a category at FTDNA?
Logged
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #11 on: December 05, 2009, 09:13:29 PM »

The UK is a category, but Great Britain is not.

Right now the 12-marker+ count for those who list UK as their place of ancestral origin is 8,766. I agree with you that most of those are probably English.
« Last Edit: December 05, 2009, 09:19:05 PM by rms2 » Logged

Jdean
Old Hand
****
Offline Offline

Posts: 678


« Reply #12 on: December 05, 2009, 09:25:29 PM »

The UK is a category, but Great Britain is not.

Not that surprising really, not that many people this side of the pond know the difference so I would expect even less in the US (where most interest in DNA genealogy resides) would either.
Logged

Y-DNA R-DF49*
MtDNA J1c2e
Kit No. 117897
Ysearch 3BMC9

rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #13 on: December 05, 2009, 09:43:38 PM »

Here are the figures as of right now for the British Isles.

12-marker+ Tests per Ancestral Origins

England -   19,703
Scotland -    9,110
Wales -        1,651
N. Ireland -     577
UK -              8,766
Ireland -     11,384
Total -         51,191


Here are some figures for some continental European countries.

Austria -       494
Belgium -      433
Croatia -       186
Czech Rep. - 508
Denmark -     697
Finland -     1,291
France -      2,764
Germany -  9,948
Italy -         2,877
Lith. -            817
Lux.                 45
Neth. -       1,384
Norway -   1,098
Poland -     2,987
Portugal -     647
Slovakia -     436
Spain -       2,792
Sweden -   1,349
Switz. -      1,458
Total -       32,166

« Last Edit: December 05, 2009, 11:18:40 PM by rms2 » Logged

NealtheRed
Old Hand
****
Offline Offline

Posts: 930


« Reply #14 on: December 06, 2009, 11:56:53 AM »

There is no doubt that some people have been blinded by numbers and ignored the sampling biase.  The thing is I think this was obvious a long time ago.  I kept my eye only on the hit Rate for L21 testing in France and ever since the first 6 or 7 tests went in a roughly 50% hit rate has been apparent.  I couldnt understand those who just looked at totals and ignored the huge sampling bias.  I can only guess it was down to the genealogically orientated being desperate to find an isles specific marker but it was a mass logic bypass.  In fact, that this was the main interest in L21 for many is shown by the striking lack of interest in L21 on rootsweb etc where it kind of feels like L21 was blanked as it has emerged more and more clearly it was not an isles-specific clade.  Meanwhile the rush to get more downstream localised SNP continues apace.  I think it is just a fact of life that most in the hobby are looking for country-specific certainty and essentially a more shallow time perspective.  Their main aim is to find ever more localised and even semi-private SNPs that indicates their yDNA origin very closely.  Problem is for those interested in deeper time is that localised SNPs do not tell us anything about migrational history.        

That's right. The current focus on downstream SNPs is kind of frustrating to me. Tests for them are being returned with lightning speed while good ol' L21 tests seem to be taking the full time predicted and more.

I got so excited by L21 (and I still am excited by it) that I must admit further downstream SNPs don't have that much appeal (unless one were found that split L21 in two or something like that). I'm still thrilled with the quest to see how L21 is distributed in Europe, and I have a hard time understanding those L21+ guys who don't share that interest.

As an aside, the L159.2 SNP that was found does give the Irish Sea group genealogical value (it's cluster-specific), but as you said, L21 is just starting to make its rounds across Europe. These new Czech results are adding to the interesting spread of L21.
Logged

Y-DNA: R-Z255 (L159.2+) - Downing (Irish Sea)


MTDNA: HV4a1 - Centrella (Avellino, Italy)


Ysearch: 4PSCK



GoldenHind
Old Hand
****
Offline Offline

Posts: 731


« Reply #15 on: December 06, 2009, 04:32:58 PM »

Excellent work, Rich.
Logged
alan trowel hands.
Guru
*****
Offline Offline

Posts: 2012


« Reply #16 on: December 06, 2009, 07:10:49 PM »

,
I guess one thing that can be said of new SNPs downstream of L21 is that they might be able to establish something of a trail that could tell us a bit of the migration story of L21.
I suppose that all depends on whether L21 did most of its spreading soon or long after the SNP occurred.  If it was soon after then there will be no non-localised SNP that splits L21 in half.  Its unclear if L21 mainly spread to its current locations in one period or not but the similar variance MRCA calculations for all R1b1b2 clades could be taken to suggest that there was some sort of huge population explosion and spread at one point in time that may account for most of it. THe local intra-clade variance calculation seemed similar in the isles and Europe.  This of course is far from clinching.  It is possible alternatively that L21 arrived in some areas much later or in waves.  All that is certain is that it arrived in the isles at some point between the L21* MRCA date and that of M222.  According to the usual number crunchersl, that could put it arriving in the isles anytime from the late Neolithic to the Iron Age.   However, I have no confidence that these calculations will not alter when various assumptions are tweaked.  
« Last Edit: December 06, 2009, 07:13:52 PM by alan trowel hands. » Logged
Mike Walsh
Guru
*****
Offline Offline

Posts: 2964


WWW
« Reply #17 on: December 07, 2009, 03:52:07 PM »

I guess one thing that can be said of new SNPs downstream of L21 is that they might be able to establish something of a trail that could tell us a bit of the migration story of L21.
This is true.  The are critical to determining direction, similar to the ht35 project as it relates to M-269 overall.
Logged

R1b-L21>L513(DF1)>S6365>L705.2(&CTS11744,CTS6621)
rms2
Board Moderator
Guru
*****
Offline Offline

Posts: 5023


« Reply #18 on: December 09, 2009, 12:27:12 PM »

Remember what I said earlier about Ancestral Origins numbers always going up?

Well, as of today, Ireland is up 17 since I last wrote, from 11,384 to 11,901, and Scotland is up 18, from 9,110 to 9,128. England is up 23, from 19,703 to 19,726.

A small fraction of that may have come from some guys changing their ancestral affiliation: "United Kingdom" is down 6, from 8,776 to 8,770.

As a whole the British Isles are up 66, from 51,191 to 51,257.

France is unchanged at 2,764, same as it was a few days ago.

« Last Edit: December 09, 2009, 12:35:32 PM by rms2 » Logged

Pages: [1] Go Up Print 
« previous next »
Jump to:  


SEO light theme by © Mustang forums. Powered by SMF 1.1.13 | SMF © 2006-2011, Simple Machines LLC

Page created in 0.121 seconds with 18 queries.