New haplogroup I2a map

No replay from Joro, that try to privatize a haplogorup. Maybe he learned about genes and come to his senses. Cheers!
 
Wiik gets an average 29.7% of I. He cites Balanovsky, Rootsi, Laitinen, Rosser and Nasidze, but doesn't give the detail of the calculation.

I checked each study, but only Rosser gives clear percentages of I for Belarus (34%) but the sample is small (n=41) and subclades aren't distinguished. It is the same study that found 42% of haplogroup I in the Bulgarians, with an even smaller sample size (only 24 individuals !).

I will lower the percentage of I2a for Belarus from 25% to 18% in the Y-DNA table. Balanovsky's average is about 18%. If it is less I don't know what other haplogroup to increase without contradicting other studies.
Here's the data about Belarus regarding I2a1b

KA4Ef3V.png
 
I disagree a bit with your percentages on the iberian peninsula. Iberian Roots with a sample size of more than 2400 has listed the following:
Navarre (Spain): 9.09%
Aragon (Spain): 5.33%
Catalonia (Spain): 2.05%
Basque Country (Spain): 1.84%
Andalucia (Spain): 0.72%
Galicia (Spain): 0.68%
Castille y Leon (Spain): 0.63%
All others (Spain) : <0.5%
Spain (Total): 1.21% (!)
Azores (Portugal): 8.16% (!)
Vila-Real (Portugal): 4.76%
Evora (Portugal): 3.70%
Viseu (Portugal): 3.13%
Santarem (Portugal): 2.98%
Leiria (Portugal): 2.78%
Braganca (Portugal): 2.56%
Aveiro (Portugal): 1.75%
Braga (Portugal): 1.52%
Lisboa (Portugal): 1.15%
All others (Portugal): <1%

Seems like I2a tends too be more on the Atlantic Side than on the Mediterranean Side of Iberia.
All data is from iberianroots dot com
 
I disagree a bit with your percentages on the iberian peninsula. Iberian Roots with a sample size of more than 2400 has listed the following:
Navarre (Spain): 9.09%
Aragon (Spain): 5.33%
Catalonia (Spain): 2.05%
Basque Country (Spain): 1.84%
Andalucia (Spain): 0.72%
Galicia (Spain): 0.68%
Castille y Leon (Spain): 0.63%
All others (Spain) : <0.5%
Spain (Total): 1.21% (!)
Azores (Portugal): 8.16% (!)
Vila-Real (Portugal): 4.76%
Evora (Portugal): 3.70%
Viseu (Portugal): 3.13%
Santarem (Portugal): 2.98%
Leiria (Portugal): 2.78%
Braganca (Portugal): 2.56%
Aveiro (Portugal): 1.75%
Braga (Portugal): 1.52%
Lisboa (Portugal): 1.15%
All others (Portugal): <1%

Seems like I2a tends too be more on the Atlantic Side than on the Mediterranean Side of Iberia.
All data is from iberianroots dot com

You are right, there is too much I2a1 in Andalusia and Galicia. I will revise that now.

However, many of the percentages listed by Iberian roots contradict those of scientific papers.

The percentage for Andalusia cannot be 0.72% when Adams et al. (n=168) found 2.5%, Flores et al. (n=258) found 3.2%, and Rootsi et al (n=103) found at least 1% (+2.9% of unidentified I*). That's a sample size of 529, against 276 for Iberian roots.

I2a1 cannot be 0.63% in Castille, when Adams + Flores have 7x I2a1 out of 213 samples (3.3%).

And finally, how could there be only 1.84% of I2a1 in the Basque country when Adams et al. (n=116) found 6.9%, Rootsi et al. (n=100) found 6%, Flores et al (n=45) found 4.4%, and Young et al. (n=126) found 2.4%. That's a total of 387 against 327 for Iberian roots.
 
Last edited:
The Instituto Nacional de Estadistica of Spain tell us
that the male population was on 01.01.2012:

23.298.356 total male in Spain

4,180,285 male in Adalucia

1,261,141 male Castilla y Leon

1,069,648 male Castilla – La Mancha

1,070,438 male Pais Vasco

etc.

Do you really think the percentages of Iberian roots and the other mentioned scientific papers are representative?

I think, the samples sizes are too few for a good statistic and it represent only one part of the population.

For a representative statistic you need a good miscellaneous of the population and more samples. For instance:

100.000 of the male Pais Vasco or so?
 
Maciamo what study shows lower then 10% frequency of I2a1 in central Ukraine?
 
The Instituto Nacional de Estadistica of Spain tell us
that the male population was on 01.01.2012:

23.298.356 total male in Spain

4,180,285 male in Adalucia

1,261,141 male Castilla y Leon

1,069,648 male Castilla – La Mancha

1,070,438 male Pais Vasco

etc.

Do you really think the percentages of Iberian roots and the other mentioned scientific papers are representative?

I think, the samples sizes are too few for a good statistic and it represent only one part of the population.

For a representative statistic you need a good miscellaneous of the population and more samples. For instance:

100.000 of the male Pais Vasco or so?



100000 Basques??? My personal experience tells me we don't need a so huge sample to have representative enough percentages - more restricted but more numerous areas studied are of better use than a huge global or to roughly regional sample - no offense -
but it is true, the rarest a type in a sample, the bigger the risk of "statistical drift"; so what would be sufficient in Iberia concerning Y-R1b scores, could be unsufficient for rarest HGs or 'subHGs'
good afternoon
 
You are right, there is too much I2a1 in Andalusia and Galicia. I will revise that now.

However, many of the percentages listed by Iberian roots contradict those of scientific papers.

The percentage for Andalusia cannot be 0.72% when Adams et al. (n=168) found 2.5%, Flores et al. (n=258) found 3.2%, and Rootsi et al (n=103) found at least 1% (+2.9% of unidentified I*). That's a sample size of 529, against 276 for Iberian roots.

I2a1 cannot be 0.63% in Castille, when Adams + Flores have 7x I2a1 out of 213 samples (3.3%).

And finally, how could there be only 1.84% of I2a1 in the Basque country when Adams et al. (n=116) found 6.9%, Rootsi et al. (n=100) found 6%, Flores et al (n=45) found 4.4%, and Young et al. (n=126) found 2.4%. That's a total of 387 against 327 for Iberian roots.

You are right in a way, but especially for Portugal it's helpful in my opinion. I'd suggest using the weighted average for Spain:
Andalusia: (n=805) : 1,92%
Castile: (n=373) : 2,15%
Basque Co.: (n=714) : 3,50 %
 
I think, it is better to test a good mix of regional samples (100.000 is always better than 387 or 529) and then add the results of the regions together.

Why should we exclude the rarest haplogroups, when we like to find out the percentages of a region? When you test 100.000 samples you have better results for the rare haplogroups and the percentage of „R“ will be a little bit or more lower.

In any case, only 387 or 327 samples are never representative for a population of 1.000.000 male, because we can never have a good mix with so few samples.

The proof is: The percentages of the scientific papers are too different and inexact (and also the Iberian roots).


The question is: How many samples do we really need for a good Y-DNA statistic in proportion to the number of males in a population?
 
Is there already any I2a1b subclade found in Kurds or it is a suggestion on the map? I know only for I2a2.
 

This thread has been viewed 179759 times.

Back
Top