PDA

View Full Version : Credibility of FTDNA



castelleone
23-07-17, 10:13
In your opinion,
how credible is FTDNA's results concerning identification of ethnic groups?

binx
24-07-17, 01:04
Low credibility

Carl Graham
25-07-17, 02:41
I'll be honest I thought my original MyOrigins was reasonably decent and in line withe the general trend of tests , but the new "upgrade " is pretty lousy.

castelleone
28-07-17, 19:18
I'll be honest I thought my original MyOrigins was reasonably decent and in line withe the general trend of tests , but the new "upgrade " is pretty lousy.

Did you get any reasonable answer from their forum? It seems 100% amateurish.

Carl Graham
29-07-17, 21:17
Their forums are so mediocre I don't bother with them.

castelleone
12-08-17, 19:37
Their forums are so mediocre I don't bother with them.

I had the same impression,
since they seem to allow for so much delusional misinformation to spread.
What bothers you the most?

murad1234
30-08-17, 13:20
I think it is better then ancestry in my opinion, as I stated I think it did a better job of break down the Balkan and Eastern Europe areas then Ancestry does.

AbdoNumen
21-11-17, 18:45
As far as Middle/Near Eastern populations go, FTDNA is one of the only companies out there brave enough to break the region down into sub-regions. Though from the samples I've seen, it is heavily weighted toward Asia Minor.

My hope is that myOrigins 3.0 will introduce other regions and modal populations such as Caucasus (Nakh, Georgian, Adyghe), Persia, Arabia, and will incorporate ancient DNA for regions that were desolate for much of modern history like southern Levant.

I1a3_Young
21-11-17, 23:08
In your opinion,
how credible is FTDNA's results concerning identification of ethnic groups?
They inflate Scandinavian percentages to unbelievable levels for me.

backdrop12
25-11-17, 17:17
Very low credibility . I have english in my Ancestry ( lee, Green , Downes,etc) and I got zero percent Great britain and irish.

mwauthy
26-11-17, 13:02
Mom(French Canadian)
West And Central Europe: 99%
East Middle East: 2%

Dad(Wallonia Belgian)
West And Central Europe: 80%
Southeast Europe: 13%
Scandinavia: 6%
Finland: 2%

Me
West And Central Europe: 90%
Southeast Europe: 10%

OkTex
26-11-17, 15:16
Pretty happy with my FTDNA calls...myself, an uncle and a double-cousin all seem accurate... I did not have pre-conceived ethnicities based on guesses or surnames.

castelleone
28-11-17, 23:01
Pretty happy with my FTDNA calls...myself, an uncle and a double-cousin all seem accurate... I did not have pre-conceived ethnicities based on guesses or surnames.

is this equivalent to say, you didnt need the test because it suffices to look at yourselves in the mirror?

Mich Glitch
29-11-17, 19:34
Show me the best company please.
Any results concerning identification of ethnic groups are low credible. That's all.

castelleone
30-11-17, 20:26
Show me the best company please.
Any results concerning identification of ethnic groups are low credible. That's all.

Thank you for enriching this thread with generalization and inductive fallacy.

mwauthy
25-12-17, 05:03
My wife is half Italian (Naples and Northern Italy). Here are her Southern Europe scores from various companies:

23andMe:
Italian: 32.5%
Iberian: 5.2%
Broadly Southern: 8.9%

DNA Land:
South/Central European: 33%
Southwestern European: 4.6%


FTDNA:
Iberian: 32%
Southeastern Europe: 14%
Sephardic: 14%

Is the FTDNA algorithm really that off or does this mean it’s just really hard to differentiate Southern European alleles?

davef
25-12-17, 06:28
In 23 and me she's 45 percent southern, DNA land she's 37 percent southern, ftdna she's 60 percent Southern. What did she get in the other categories (each test listed)?

mwauthy
25-12-17, 21:31
In 23 and me she's 45 percent southern, DNA land she's 37 percent southern, ftdna she's 60 percent Southern. What did she get in the other categories (each test listed)?

To be more precise my wife is half Italian (Naples/Amalfi and Northern Italy), and half American (mostly British Ancestry). Full scores listed below with the addition of Myheritage.

23andMe:
Southern European 46.5%
Italian: 32.5%
Iberian: 5.2%
Broadly Southern: 8.9%
Northwestern European: 43.6%
British and Irish: 20.9%
French and German: 6.3%
Broadly Northwestern: 16.5%
Broadly European: 9%
Native American: 0.5%


DNA Land:
Northwest European: 41%
South/Central European: 33%
Ashkenazi: 8%
Central Indoeuropean: 4.8%
Southwestern European: 4.6%
North African 4.4%
Native American: 1.8%
Arab/Egyptian: 1.1%


Myheritage:
Italian: 59.3%
English: 17.3%
North and West European: 10.7%
Irish, Scottish, And Welsh: 6.1%
Ashkenazi Jewish: 5.6%
Native American: 1%


FTDNA:
British Isles: 33%
Iberian: 32%
Southeastern Europe: 14%
Sephardic: 14%
West and Central Europe: 3%
South Central Asia: <2%
Asia Minor: <2%
Ashkenazi:<1%
Oceania:<1%
East Europe: <1%
North and Central America: <1%

Angela
26-12-17, 01:47
To be more precise my wife is half Italian (Naples/Amalfi and Northern Italy), and half American (mostly British Ancestry). Full scores listed below with the addition of Myheritage.

23andMe:
Southern European 46.5%
Italian: 32.5%
Iberian: 5.2%
Broadly Southern: 8.9%
Northwestern European: 43.6%
British and Irish: 20.9%
French and German: 6.3%
Broadly Northwestern: 16.5%
Broadly European: 9%
Native American: 0.5%


DNA Land:
Northwest European: 41%
South/Central European: 33%
Ashkenazi: 8%
Central Indoeuropean: 4.8%
Southwestern European: 4.6%
North African 4.4%
Native American: 1.8%
Arab/Egyptian: 1.1%


Myheritage:
Italian: 59.3%
English: 17.3%
North and West European: 10.7%
Irish, Scottish, And Welsh: 6.1%
Ashkenazi Jewish: 5.6%
Native American: 1%


FTDNA:
British Isles: 33%
Iberian: 32%
Southeastern Europe: 14%
Sephardic: 14%
West and Central Europe: 3%
South Central Asia: <2%
Asia Minor: <2%
Ashkenazi:<1%
Oceania:<1%
East Europe: <1%
North and Central America: <1%

This is what I mean when I say that I don't trust FTDNA at all.

Let's assume that someone was adopted, and was half Italian. FTDNA would steer this person totally in the wrong direction.

My Heritage isn't bad, but it overstates the percentage. DNA Land understates it, and uses the term "South Central European". Do they tell people that means Italian?

23andme needs one little bit of explanation, which is that if someone gets a large percentage of Italian and then minority bits of Iberian and Southern European, they should add all of that together, because that's very close to the percentage. Is this the v5 result? If it isn't, v5 might get even closer. It did for me.

mwauthy
26-12-17, 03:49
This is what I mean when I say that I don't trust FTDNA at all.
Let's assume that someone was adopted, and was half Italian. FTDNA would steer this person totally in the wrong direction.
My Heritage isn't bad, but it overstates the percentage. DNA Land understates it, and uses the term "South Central European". Do they tell people that means Italian?
23andme needs one little bit of explanation, which is that if someone gets a large percentage of Italian and then minority bits of Iberian and Southern European, they should add all of that together, because that's very close to the percentage. Is this the v5 result? If it isn't, v5 might get even closer. It did for me.
I agree with your analysis. It was 23andMe V4.
They do mention on DNA Land that South Central European is for central/northern Italy because the samples they use are from Tuscany and Bergamo.

davef
26-12-17, 06:18
How much % north Italian and % south Italian is she?

Oh and cute kitty avatar Angela :)

mwauthy
26-12-17, 18:42
How much % north Italian and % south Italian is she?

Oh and cute kitty avatar Angela :)

Her Grandma is from Naples/Amalfi area and her Grandpa is from Northern Italy. So 1/4 Northern. It depends on whether or not you consider Naples closer to central Italy or Southern Italy for the other 1/4. On DNA Land she did not get any Mediterranean Islander which has samples from Sicily or any Sardinian. If you add her South Central European 33% with Southwestern 4.6% with Central IndoEuropean 4.8% with North African 4.4% with Arab/Egyptian 1.1% you get 47.9% or almost half Italian. This is also close to 46.5% Southern Europe score from 23andMe.

davef
26-12-17, 19:07
Her Grandma is from Naples/Amalfi area and her Grandpa is from Northern Italy. So 1/4 Northern. It depends on whether or not you consider Naples closer to central Italy or Southern Italy for the other 1/4. On DNA Land she did not get any Mediterranean Islander which has samples from Sicily or any Sardinian. If you add her South Central European 33% with Southwestern 4.6% with Central IndoEuropean 4.8% with North African 4.4% with Arab/Egyptian 1.1% you get 47.9% or almost half Italian. This is also close to 46.5% Southern Europe score from 23andMe.

Naples is without a doubt in southern Italy, so that makes her 1/4 southern Italian.

Amerijoe
26-12-17, 20:19
Didn’t vote due to lack of paternal info., but here are the results and some comparisons. Maternal side paperwork traces back through Scot/Irish Ancestry to around mid 1700’s. Family surnames are mentioned as far back as Dal Riada, so I should have a high degree of British. FTDNA breakdown is at 98% British Isles and 2% Ashkenazi. Ancestry 49% I/S/W, 47% GB, 1% European Jewish. 23andMe 90.8% B&I, Scandinavian 1.4%, Broadly NW Eur. 6%. Geno2 B&I 97%, 2% Arabian. Other tests taken all fall within the same percentage framework. Not a great amount of variation for me between the various testing companies.

mwauthy
26-12-17, 23:21
Didn’t vote due to lack of paternal info., but here are the results and some comparisons. Maternal side paperwork traces back through Scot/Irish Ancestry to around mid 1700’s. Family surnames are mentioned as far back as Dal Riada, so I should have a high degree of British. FTDNA breakdown is at 98% British Isles and 2% Ashkenazi. Ancestry 49% I/S/W, 47% GB, 1% European Jewish. 23andMe 90.8% B&I, Scandinavian 1.4%, Broadly NW Eur. 6%. Geno2 B&I 97%, 2% Arabian. Other tests taken all fall within the same percentage framework. Not a great amount of variation for me between the various testing companies.
What are your DNA Land scores?

Pax Augusta
26-12-17, 23:44
Her Grandma is from Naples/Amalfi area and her Grandpa is from Northern Italy. So 1/4 Northern. It depends on whether or not you consider Naples closer to central Italy or Southern Italy for the other 1/4.

As Davef said, there is no doubt that Naples/Amalfi is in southern Italy. From which area of northern Italy the other 1/4?

Amerijoe
27-12-17, 06:10
What are your DNA Land scores?

DNA.LAND
Northwest European 95%, South/Central Europe 2.7%, Ambiguous 2.2%.

LivingDNA
GB&I 94.3%, Scandinavia 4.6%, Kurdish 1.1%. This is the highest Scandinavian score of all tested companies to date which is surprising since most other tests show 0%. You would think being a Scot, Scandinavian would show across all testing platforms.

mwauthy
27-12-17, 15:06
As Davef said, there is no doubt that Naples/Amalfi is in southern Italy. From which area of northern Italy the other 1/4?

I’m not aware of the exact location in Nothern Italy.

mwauthy
27-12-17, 15:16
DNA.LAND
Northwest European 95%, South/Central Europe 2.7%, Ambiguous 2.2%.

LivingDNA
GB&I 94.3%, Scandinavia 4.6%, Kurdish 1.1%. This is the highest Scandinavian score of all tested companies to date which is surprising since most other tests show 0%. You would think being a Scot, Scandinavian would show across all testing platforms.

There are a lot of overlapping alleles between Northwestern Europe, Great Britain, and Scandinavia because of cross migrations over millennia. Just because you match closer to modern UK reference samples does not mean that there is not a lot of Scandinavian influence in the UK or in your genome since the Nordic Bronze Age. It’s not one or the other when it comes to these percentages.

Amerijoe
27-12-17, 17:28
There are a lot of overlapping alleles between Northwestern Europe, Great Britain, and Scandinavia because of cross migrations over millennia. Just because you match closer to modern UK reference samples does not mean that there is not a lot of Scandinavian influence in the UK or in your genome since the Nordic Bronze Age. It’s not one or the other when it comes to these percentages.

I am under no allusion of the interpretive nature of commercial genetic testing. Just that the inconsistencies in that interpretation. I was schooled and worked in an area related to the physical sciences where computation and formula interpretation were exacting. Ancestry testing by many of these companies borders on novelty science at best.

When I first decided to pursue my father’s origin, I assumed he fell within one of the Scottish clans. Not only did my dna not fall within a Scot clan, my ydna didn’t even fall within the British Isles. The dna has a great affinity for NW Europe in which Scandinavia falls.

Signed up to join FTDNA Scottish Project. Listed me as not Scottish but possibly Scandinavian. Joined Scandinavian Project, listed as ungrouped. Accepted the fact even though being native born, my dna is an outlier with no matches within recent history. The two recent BigY matches seem to have brought me back to the Isles. Even though it’s 600 yrs. it gives me hope, more may materialize.

Note, using my maternal aunt as a proxie for my mum, her Ancestry Scandinavian is 10%, mine is 0%. To conclude, I really don’t care what labels are used in my dna interpretation, as long as the underlying data is correct. Hopefully, one day these science may make these interpretation discussions irrelevant.

mwauthy
27-12-17, 22:45
There's that mandatory Sephardic score from ftdna

Same percentage of Sephardic as well at 14% as my wife. Coincidence?

davef
27-12-17, 23:33
Same percentage of Sephardic as well at 14% as my wife. Coincidence?

They're (Sephardics) European Jews with heavy Southern European like genes (Ashkenazi in same boat in that regard) plus ftdna gravitating everyone towards them as we see in this thread. This issue needs to be cleaned up or people will get confused.

mwauthy
28-12-17, 00:47
They're (Sephardics) European Jews with heavy Southern European like genes (Ashkenazi in same boat in that regard) plus ftdna gravitating everyone towards them as we see in this thread. This issue needs to be cleaned up or people will get confused.

When Kurd created his k25 and k29 algorithm he left out those two components because some samples were getting false positives. Doesn’t make sense to me to associate genetic alleles with ethnicity or religion. It can be very misleading. Perhaps in the future companies will release allele reports and then you can see how common certain alleles you have are around the world or with ancient samples.