Middle Eastern?

[Dna29]

Hi,
I started out at 23andme. Both services vary greatly in predicting my ancestry. On one hand FT gets it right with the Scandinavian ancestry at 37% this even seems low to me but 23andme predicts Scandinavian at 8%. 23andme predicts 99.6% European which I believe is right but FT predicts only 80% European and 20% Middle Eastern which can't be right as far as I know. Why are there such large variances between the two sevices?
Thanks for any replies.

[Armando]

The difference is the methodology and the source populations or data set they chose to use. FTDNA didn't use some of the populations that 23andme did. They both used some of the same populations from Human Genome Diversity Project and so on but FTDNA didn't use all of them.

FTDNA White Paper is at https://www.familytreedna.com/learn/...s-methodology/

Reference populations:

• GeneByGene DNA customer database
• Human Genome Diversity Project
• International HapMap Project
• Estonian Biocentre


23andme Methodology

https://www.23andme.com/ancestry_composition_guide/

The public reference datasets we've drawn from include the Human Genome Diversity Project, HapMap, and the 1000 Genomes project. We perform the same filtering on these public reference datasets as we do on the customer dataset.

Your 23andme 99.6% European looks to be a much better result than your FTDNA 80% European and 20% Middle Eastern. Across the board people report better results with 23andme than with FTDNA myOrigins.

[Germanica]

Hi,
I started out at 23andme. Both services vary greatly in predicting my ancestry. On one hand FT gets it right with the Scandinavian ancestry at 37% this even seems low to me but 23andme predicts Scandinavian at 8%. 23andme predicts 99.6% European which I believe is right but FT predicts only 80% European and 20% Middle Eastern which can't be right as far as I know. Why are there such large variances between the two sevices?
Thanks for any replies.

Do you have any Mediterranean ancestry? That often comes up as Middle Eastern - for example, one of my grandparents was Italian and FTDNA says I'm 17% Middle Eastern. If you add that to my 20% Southern European results, that's about 37%, which is very close to AncestryDNA's estimate of 31% Italy/Greece.

23andMe have "broadly" categories, like "Broadly Northern European" or "Broadly European"- this is basically them saying "we can't narrow this amount of DNA down any further than this", which is honestly more accurate than other companies. Do you have any "Broadly" results from 23andMe? Because if you do, that could be why your Scandinavian results are so low with them - a lot of your Scandinavian DNA could be too similar to other groups to be able to tell it apart, therefore it's getting put into a "Broadly" category.

Without knowing more about your family tree on paper and your full DNA ethnicity results, it's difficult to say why exactly the results vary.

[Dna29]

Thanks for the replies. They both shed more light on the situation. Yes, I am part Italian and yes I do have 25.6% broadly Northern European from 23andme. I also have 10.9% British and Irish but from where I have no idea. Anyway thanks for answering my post. I think that FT should not be trying to give ancestry backgrounds when it is so far off, but then again they got all of the Scandinavian when 23andme could not.

[dna]

@Dna29

Some people just want to have some idea. Let's say, my skin is black. Am I from India or from Ethiopia? A Melanesian?

You know your ancestors from past 500 years. myOrigins gives you an educated guess who possibly your ancestors were some 1000-10000 years ago. I am mildly surprised that it can predict anything :-)

The bottom line is that your deep ancestry can have very different percentages than you could guess based on your ancestors from 500 years ago.

W.

[Armando]

Some people just want to have some idea. Let's say, my skin is black. Am I from India or from Ethiopia? A Melanesian?

You know your ancestors from past 500 years. myOrigins gives you an educated guess who possibly your ancestors were some 1000-10000 years ago. I am mildly surprised that it can predict anything :-)

The bottom line is that your deep ancestry can have very different percentages than you could guess based on your ancestors from 500 years ago.

W.

The bottom line is 23andme does a better job. FTDNA needs to fix myOrigins. If other companies do a better job then so can FTDNA. We shouldn't have to settle for sub par results. A person with partial Italian ancestry shouldn't be getting 20% Middle Eastern even if Italians have 20% of their DNA is from the Middle East 8,000 years ago because the competition doesn't do that.

[Armando]

Thanks for the replies. They both shed more light on the situation. Yes, I am part Italian and yes I do have 25.6% broadly Northern European from 23andme. I also have 10.9% British and Irish but from where I have no idea. Anyway thanks for answering my post. I think that FT should not be trying to give ancestry backgrounds when it is so far off, but then again they got all of the Scandinavian when 23andme could not.

In my opinion it's much better to be conservative and provide 25.6% broadly Northern European for Scandinavian than to be completely wrong and provide 20% Middle Eastern for Italian ancestry. Both Scandinavia and Italy are European countries so 23andme isn't providing an incorrect result. Italy is not a Middle Eastern country so FTDNA is providing an incorrect result.

[Germanica]

In my opinion it's much better to be conservative and provide 25.6% broadly Northern European for Scandinavian than to be completely wrong and provide 20% Middle Eastern for Italian ancestry. Both Scandinavia and Italy are European countries so 23andme isn't providing an incorrect result. Italy is not a Middle Eastern country so FTDNA is providing an incorrect result.

It's not really incorrect since Italians, especially southern Italians and Sicilians, do have a lot of Middle Eastern genetic influences. I find it really fascinating that there is an ancient Middle Eastern component of my Italian heritage that I wouldn't have otherwise known about were it not for FTDNA.

[Táltos]

It's not really incorrect since Italians, especially southern Italians and Sicilians, do have a lot of Middle Eastern genetic influences. I find it really fascinating that there is an ancient Middle Eastern component of my Italian heritage that I wouldn't have otherwise known about were it not for FTDNA.

If FTDNA really wanted to give us a unique product here, they should give us settings to go from say 1,000 yrs ago to only the past 500 years. Then you could go back and forth with the settings to compare.

I have to agree with Armando that they dropped the ball here with their update. My mom who is half Southern Italian of Albanian descent scores 23% Middle Eastern. 23andme assigns her only 0.1% Middle Eastern. They also give her 17.9% Balkan, 12.5% Italian, and 18.2% Broadly Southern European. For me that is more informative than the 23% Middle Eastern.

[dna]

Pure curiosity...

Would you gals and guys feel better if the cluster was named Mediterranean ?

W.

[AFH]

Would you gals and guys feel better if the cluster was named Mediterranean ?

W.

I would not I like having a the cluster named Mediterranean. I prefer Middle Eastern. Being of Portuguese ancestry I am well aware of the the Middle Eastern admixture in Iberians. Berbers, Arabs and Jews for centuries lived in the Iberian penisula and all contributed to my DNA. I want to claim all of my DNA.

There is a long history in my paricular ethnic group of some trying to distance themselves from that fact. There are reasons for that.

First the role the inquisition played in Portugal, Spain and their colonies for three centuries.

Second some believe they will be better accepted by Anglo Americans by hiding that fact.

I realize everyone has a different view and experience but for me it is a matter of claiming a part of my history and heritage.

[Germanica]

If FTDNA really wanted to give us a unique product here, they should give us settings to go from say 1,000 yrs ago to only the past 500 years. Then you could go back and forth with the settings to compare.

I have to agree with Armando that they dropped the ball here with their update. My mom who is half Southern Italian of Albanian descent scores 23% Middle Eastern. 23andme assigns her only 0.1% Middle Eastern. They also give her 17.9% Balkan, 12.5% Italian, and 18.2% Broadly Southern European. For me that is more informative than the 23% Middle Eastern.

Well, there is a break down within the Middle Eastern results.

[Armando]

It's not really incorrect since Italians, especially southern Italians and Sicilians, do have a lot of Middle Eastern genetic influences. I find it really fascinating that there is an ancient Middle Eastern component of my Italian heritage that I wouldn't have otherwise known about were it not for FTDNA.

However, Italians aren't fully Middle Eastern. If Dna29 were 100% Italian and showed 20% Middle Eastern I wouldn't be complaining because 20% Middle Eastern does show up in Italians at 23andme. A person that is mostly north European and partially Italian shouldn't show 20% Middle Eastern because they really aren't.

[Armando]

If FTDNA really wanted to give us a unique product here, they should give us settings to go from say 1,000 yrs ago to only the past 500 years. Then you could go back and forth with the settings to compare.

It is extremely hard to date admixture. It seems to me that the dates provided for the myOrigins results are based on what is showing up in the results and comparing it to known migrations and not a result based on dates.

I have to agree with Armando that they dropped the ball here with their update. My mom who is half Southern Italian of Albanian descent scores 23% Middle Eastern. 23andme assigns her only 0.1% Middle Eastern. They also give her 17.9% Balkan, 12.5% Italian, and 18.2% Broadly Southern European. For me that is more informative than the 23% Middle Eastern.

Thank you.

[Armando]

Well, there is a break down within the Middle Eastern results.

The problem is that the Middle Eastern is exaggerated no matter how it is broken down.