PDA

View Full Version : Erik's Known Heritage vs. AncestryDNA vs. FTDNA



Erik
09-21-2014, 03:25 PM
Hi everyone,

I would like to show you the difference between my family history, AncestryDNA, and FTDNA.

[REDACTED]

Salkin
09-21-2014, 03:34 PM
Not bad. I'm waiting for my FTDNA kit to get here for Y-STR, mtFullSequence and Family Finder processing.

AncestryDNA looks kind of neat, but it doesn't seem compelling enough for me to try and get around their US-only policy, much as I'm sure that could be done pretty easily by recruiting a USian proxy.

Erik
09-21-2014, 03:43 PM
Not bad. I'm waiting for my FTDNA kit to get here for Y-STR, mtFullSequence and Family Finder processing.

AncestryDNA looks kind of neat, but it doesn't seem compelling enough for me to try and get around their US-only policy, much as I'm sure that could be done pretty easily by recruiting a USian proxy.
Nice, you should also make a thread once you get your results.

It's not worth it if you're outside of the USA, to be honest. FTDNA and Genographic is what I'm recommending for my cousins in Sweden.

Salkin
09-21-2014, 04:05 PM
Makes sense. I tested with 23andMe at first. They seem to have the most well-rounded offering (especially for the price), but I know their Y-DNA G haplotree is outdated and possibly dodgy, plus their v4 chip tests fewer SNPs than Family Finder, etc. I also like the look of FTDNA's project model quite a lot.

I will make a thread, or post to an existing one that seems to be about myOrigins etc.

vettor
09-21-2014, 06:19 PM
Makes sense. I tested with 23andMe at first. They seem to have the most well-rounded offering (especially for the price), but I know their Y-DNA G haplotree is outdated and possibly dodgy, plus their v4 chip tests fewer SNPs than Family Finder, etc. I also like the look of FTDNA's project model quite a lot.

I will make a thread, or post to an existing one that seems to be about myOrigins etc.

I think Natgeno2 is the best way to go for you .................you can transfer the results to ftdna for free ...............but from April 2014 , ftdna refuse to accept the negative ones from Natgeno2.

But, you can download your natgeno2 results and compare them to Isogg index chart and find these negatives, which in the end will save you money in testing SNP's ...............in the end you get a ftdna project site and a look at your negatives at your leisure

Its a great pity Ftdna stopped accepting the natgeno2 negative results...............seems like it must be something to do with stuffing up ftdna new haplotrees

All my presumed positives in ftdna tree are all negative SNP's for me

zulkif
09-21-2014, 06:43 PM
All my presumed positives in ftdna tree are all negative SNP's for me

Wow, really, how come... I see your Y-DNA is T1 - is this result from GENO 2.0? What were the presumed positives at FTDNA?

Salkin
09-21-2014, 06:57 PM
I think Natgeno2 is the best way to go for you .................you can transfer the results to ftdna for free ...............but from April 2014 , ftdna refuse to accept the negative ones from Natgeno2.

But, you can download your natgeno2 results and compare them to Isogg index chart and find these negatives, which in the end will save you money in testing SNP's ...............in the end you get a ftdna project site and a look at your negatives at your leisure

Its a great pity Ftdna stopped accepting the natgeno2 negative results...............seems like it must be something to do with stuffing up ftdna new haplotrees

All my presumed positives in ftdna tree are all negative SNP's for me

Hmm, perhaps I ought to order a NatGeno 2 test as well. I guess it'd be interesting to see my Denisovan percentage, though it's most likely close to zilch. My FTDNA kit is already underway to me. NatGeno seems to test very few SNPs according to the chart I found (http://www.isogg.org/wiki/Autosomal_DNA_testing_comparison_chart), but perhaps they have changed since.

I don't like the sound of Genographic transfers no longer working for FTDNA, though, so maybe I should hold off a little.

Thanks for the info though.

I'm figuring that a) my Y-111 STR test will probably show a great deal of haplogroup info, and FTDNA's haplotrees are way better than 23andMe's, dunno about NatGeno.
and b) If I do turn out to require additional investigation, it's possible project admins could scrounge up discount coupons or outright sponsorship for the Big Y test. As a Swede with Y-DNA G, it's fairly possible I'd be regarded as an interesting research subject.

vettor
09-21-2014, 07:01 PM
Hmm, perhaps I ought to order a NatGeno 2 test as well. I guess it'd be interesting to see my Denisovan percentage, though it's most likely close to zilch. My FTDNA kit is already underway to me. NatGeno seems to test very few SNPs according to the chart I found (http://www.isogg.org/wiki/Autosomal_DNA_testing_comparison_chart), but perhaps they have changed since.

I don't like the sound of Genographic transfers no longer working for FTDNA, though, so maybe I should hold off a little.

Thanks for the info though.

I'm figuring that a) my Y-111 STR test will probably show a great deal of haplogroup info, and FTDNA's haplotrees are way better than 23andMe's, dunno about NatGeno.
and b) If I do turn out to require additional investigation, it's possible project admins could scrounge up discount coupons or outright sponsorship for the Big Y test. As a Swede with Y-DNA G, it's fairly possible I'd be regarded as an interesting research subject.


the transfers do work , but ftdna only accept the positive ones

vettor
09-21-2014, 07:08 PM
Wow, really, how come... I see your Y-DNA is T1 - is this result from GENO 2.0? What were the presumed positives at FTDNA?

I test 67 markers with ftdna first.............a year later i got a gift for natgeno2.......so I tested with it .....and transferred my results.
My ftdna was upgraded once natgeno2 results came in ..................but it lacked the negatives

I then downloaded my natgeno2 results and compared them with Isogg index chart.......and found that all my presumes postive markers in ftdna are ALL negative markers.

In the end natgeno2 gave me more results than ftdna did.................but ftdna is better for continuation of your haplogroup

I now have a few SNP to test in ftdna .......but with my natgeno2 results.....I see clearly they will be negative for me ...............so I have reached the end of the line and must wait on Isogg to find more subclades ...............they have found 6 IIRC which are downstream from my L446 ( T1a2b )

Salkin
09-21-2014, 07:09 PM
the transfers do work , but ftdna only accept the positive ones

Noted. I'd like to have a full list of good/bad/neutral SNPs. I gather with my 23andMe plus FTDNA data pooled, once they can be obtained I should get a more comprehensive Promethease report.

Erik
09-21-2014, 07:44 PM
Does anyone think that I should get 23andme? It could either give me a better estimate or just more data for comparative purposes.

Speaking of comparing things, I'm working on something right now that's sort of interesting. I will post it soon.

Scarlet Ibis
09-21-2014, 09:35 PM
I decided to transfer my raw data to FTDNA and try to see what their results would be.
They seem more accurate.....


Thanks for sharing. I had the same experience with AncestryDNA. The ancestry results were a bit of a fail. The "Asia East" portion is ok, but the description is too broad to be wrong. As for the European portion, it gave me 22% Scandinavian, and 11% Iberian. I have no known Iberian ancestry, and I don't have that much Scandinavian ancestry. Not even close. I transferred my results to FTDNA, though, and the results were pretty decent.

However, I'm still very happy with AncestryDNA, because I didn't buy it for the ancestry composition portion of it. Bloggers, and Gedmatch are doing way more for us, in that regard, than any of the commercial testing companies are. I bought it for: 1) the raw DNA, 2) integration with my family tree on ancestry.com, and 2) so I could connect with colonial US relatives.

AncestryDNA:
http://i800.photobucket.com/albums/yy290/guesspics_bucket/myancdna1_zps144de18b.jpg

FTDNA MyOrigins...same raw data, but more accurate:
http://i800.photobucket.com/albums/yy290/guesspics_bucket/ftdna_zpsbdac4a85.jpg

Erik
09-21-2014, 10:43 PM
Thanks for sharing. I had the same experience with AncestryDNA. The ancestry results were a bit of a fail. The "Asia East" portion is ok, but the description is too broad to be wrong. As for the European portion, it gave me 22% Scandinavian, and 11% Iberian. I have no known Iberian ancestry, and I don't have that much Scandinavian ancestry. Not even close. I transferred my results to FTDNA, though, and the results were pretty decent.

However, I'm still very happy with AncestryDNA, because I didn't buy it for the ancestry composition portion of it. Bloggers, and Gedmatch are doing way more for us, in that regard, than any of the commercial testing companies are. I bought it for: 1) the raw DNA, 2) integration with my family tree on ancestry.com, and 2) so I could connect with colonial US relatives.

AncestryDNA:
http://i800.photobucket.com/albums/yy290/guesspics_bucket/myancdna1_zps144de18b.jpg

FTDNA MyOrigins...same raw data, but more accurate:
http://i800.photobucket.com/albums/yy290/guesspics_bucket/ftdna_zpsbdac4a85.jpg
Thanks for sharing your results.

I did Ancestry mainly because I wanted the ethnicity info, however I was planning on getting Genographic. The reason I chose Ancestry over Genographic was because of integration into the family tree like you said and also the cheaper price. I didn't know about FTDNA or 23andme before I got my results. I only found out about them through Google searches about raw data and how I can use it.

Speaking of GEDmatch and blogger's ethnicity calculators, which do you recommend the most? I haven't had too much success with them.

rod
09-21-2014, 11:07 PM
AncestryDNA
2653

ftDNA
2654

23andme Split View
2655

23andme Chromosome View
2656

According to 23andme I have 0.1% South Asian. The chromosome view shows the South Asian on my X chromosome, so I inherited it from my mother whose X chromosome shows 100% European.

Erik
09-21-2014, 11:09 PM
AncestryDNA
2653

ftDNA
2654

23andme Split View
2655

23andme Chromosome View
2656

According to 23andme I have 0.1% South Asian. The chromosome view shows the South Asian on my X chromosome, so I inherited it from my mother whose X chromosome shows 100% European.
Thank you for the much needed comparison between all 3. What is your known heritage?

rod
09-21-2014, 11:16 PM
Thank you for the much needed comparison between all 3. What is your known heritage?
~1/2 England, 1/8 Scotland, 1/8 Ireland, 1/8 NW Europe (France, Flanders, Netherlands, Germany), 1/8 ?

Salkin
09-22-2014, 08:27 AM
Does anyone think that I should get 23andme? It could either give me a better estimate or just more data for comparative purposes.


I really like their Ancestry Composition. Allows you to get a detailed look, including chromosome painting, and it seems to find a lot of obscure but valid stuff other calculators can't detect. (Of course, it doesn't work out for everyone... but ethnicity-by-DNA is probably always going to be more of an art than a science.)

I think AC is the most compelling reason to test with 23andMe, since they're no longer doing health reports, and you can already get Promethease with FTDNA raw data. rod's screenshot of the Split View pretty much says it all, though of course that requires getting one or both parents tested.

As for GEDmatch calculators, I've found the new MDLP K23b calculator very accurate. However, I'm a pretty easy calculator target since I have only a very small amount of admixture (though it actually suggests a Belgian component in mixed-mode! Also impressive). I think admixed people can get a little more confusing results, though I've seen some successes with K23b for them too.

Salkin
09-22-2014, 08:31 AM
Here (http://www.anthrogenica.com/showthread.php?3159-Post-your-MDLP-K23b-calculator-Oracle-ancestry-here&p=52367&viewfull=1#post52367)<-- is what the K23b Oracle-4 says about me.

Salkin
09-22-2014, 12:34 PM
Oh, and my 23andMe Ancestry Composition (http://www.anthrogenica.com/showthread.php?517-23andMe-Ancestry-Composition-Results&p=50386&viewfull=1#post50386).

Also, I think a lot of the GEDmatch calculators are primarily designed for 23andMe data, or so I've heard. You CAN run them on FTDNA data, but they might not give very good results.

MitchellSince1893
09-22-2014, 01:06 PM
Oh, and my 23andMe Ancestry Composition (http://www.anthrogenica.com/showthread.php?517-23andMe-Ancestry-Composition-Results&p=50386&viewfull=1#post50386).

Also, I think a lot of the GEDmatch calculators are primarily designed for 23andMe data, or so I've heard. You CAN run them on FTDNA data, but they might not give very good results. I see very little difference between my ftdna kit and my 23andMe kit...typically .~05% on the admixture calculators.

vettor
09-22-2014, 06:57 PM
Oh, and my 23andMe Ancestry Composition (http://www.anthrogenica.com/showthread.php?517-23andMe-Ancestry-Composition-Results&p=50386&viewfull=1#post50386).

Also, I think a lot of the GEDmatch calculators are primarily designed for 23andMe data, or so I've heard. You CAN run them on FTDNA data, but they might not give very good results.

unsure about this..........i have both kits in gedmatch and the top 3 matches for my 23andme kit is not to be seen anywhere in my ftdna matches kit ( i refer to many matches program )
the weird issue is that these top 3 have no T ( ydna ) or H ( mtdna ) ..........usually I and X respectively

I use a minimum of 9cm in both for the tests

MitchellSince1893
09-23-2014, 01:04 AM
Here's an example using MDLP K23b. My previous estimate of .05% difference is obviously off for this particular tool where it ranges from .1 to 1.12% difference; but I think you would agree there's not a whole lot of difference in the ancestral population percentages.
FTDNA
# Population Percent
1 European_Hunters_Gatherers 35.86
2 European_Early_Farmers 27.75
3 Caucasian 22.06
4 South_Central_Asian 6.07
5 Ancestral_Altaic 3.46
6 Near_East 1.71

Least-squares method.
Using 1 population approximation:
1 English @ 3.021939
2 Belgian @ 3.023029
3 Irish @ 3.023781
4 Frisian @ 3.327971
5 South_German @ 4.007581
6 North_European @ 4.257988
7 German-Volga @ 4.698667
8 English_Kent_GBR @ 4.754650
9 Dutch @ 4.987072
10 CEU @ 5.240406
11 English_Cornwall_GBR @ 5.519495
12 Welsh @ 5.751094
13 British @ 6.074735
14 Scottish_Argyll_Bute_GBR @ 7.068443
15 Orcadian @ 7.986778
16 Norwegian_East @ 8.093551
17 North_German @ 8.654167
18 Norwegian_West @ 8.682924
19 Icelandic @ 8.741966
20 Dane @ 9.199669

Using 2 populations approximation:
1 50% Irish +50% South_German @ 1.837212


Using 3 populations approximation:
1 50% CEU +25% Macedonian +25% Scottish_Argyll_Bute_GBR @ 1.242882


Using 4 populations approximation:
1 English_Kent_GBR + Macedonian + Scottish_Argyll_Bute_GBR + Scottish_Argyll_Bute_GBR @ 1.202430
2 CEU + CEU + Macedonian + Scottish_Argyll_Bute_GBR @ 1.242882
3 Irish + Montenegrian + Orcadian + Welsh @ 1.248806
4 CEU + Montenegrian + Scottish_Argyll_Bute_GBR + Welsh @ 1.268079
5 Irish + Montenegrian + Scottish_Argyll_Bute_GBR + Welsh @ 1.277259
6 Bulgarian + Icelandic + Welsh + Welsh @ 1.294320
7 CEU + Macedonian + Scottish_Argyll_Bute_GBR + Welsh @ 1.295645
8 German-Volga + Irish + South_German + Welsh @ 1.304359
9 British + CEU + Macedonian + Scottish_Argyll_Bute_GBR @ 1.323237
10 French + Irish + Scottish_Argyll_Bute_GBR + Serb_BH @ 1.328484
11 Macedonian + Scottish_Argyll_Bute_GBR + Scottish_Argyll_Bute_GBR + Welsh @ 1.336791
12 CEU + Macedonian + Scottish_Argyll_Bute_GBR + Scottish_Argyll_Bute_GBR @ 1.347564
13 Montenegrian + Scottish_Argyll_Bute_GBR + Welsh + Welsh @ 1.354631
14 English_Kent_GBR + Montenegrian + Scottish_Argyll_Bute_GBR + Welsh @ 1.360768
15 English_Kent_GBR + Macedonian + Orcadian + Scottish_Argyll_Bute_GBR @ 1.362426
16 Montenegrian + North_European + Scottish_Argyll_Bute_GBR + Welsh @ 1.365545
17 Serb_BH + Welsh + Welsh + Welsh @ 1.367487
18 Bosnian + French + Irish + Scottish_Argyll_Bute_GBR @ 1.370581
19 French + Hungarian_Budapest + Irish + Welsh @ 1.371648
20 Italian_Bergamo + Scottish_Argyll_Bute_GBR + Scottish_Argyll_Bute_GBR + Slovak @ 1.377003

23andme
# Population Percent
1 European_Hunters_Gatherers 36.50
2 European_Early_Farmers 26.53
3 Caucasian 23.18
4 South_Central_Asian 5.45
5 Ancestral_Altaic 4.30
6 Near_East 1.81

Least-squares method.
Using 1 population approximation:
1 Belgian @ 3.014118
2 South_German @ 3.096100
3 Frisian @ 3.383443
4 German-Volga @ 3.631399
5 Dutch @ 3.899249
6 English @ 3.998207
7 Irish @ 4.105394
8 North_European @ 5.204072
9 English_Kent_GBR @ 6.027362
10 CEU @ 6.612281
11 English_Cornwall_GBR @ 6.927038
12 British @ 7.328123
13 Welsh @ 7.403316
14 North_German @ 7.630266
15 Norwegian_East @ 7.838763
16 Dane @ 8.209379
17 Scottish_Argyll_Bute_GBR @ 8.341587
18 Austrian @ 8.429803
19 Norwegian_West @ 8.782618
20 Icelandic @ 9.036837

Using 2 populations approximation:
1 50% Frisian +50% German-Volga @ 1.749547


Using 3 populations approximation:
1 50% British +25% Dutch +25% Montenegrian @ 1.073732


Using 4 populations approximation:
1 English_Cornwall_GBR + Irish + Macedonian + Norwegian_West @ 0.968699
2 British + Irish + Macedonian + Norwegian_West @ 0.976407
3 English_Cornwall_GBR + Frisian + Icelandic + Macedonian @ 1.001949
4 CEU + Irish + Macedonian + Norwegian_West @ 1.010602
5 Frisian + Icelandic + Macedonian + Welsh @ 1.013239
6 Irish + Macedonian + Norwegian_West + Welsh @ 1.047157
7 Frisian + Macedonian + Norwegian_West + Welsh @ 1.057635
8 British + Frisian + Icelandic + Macedonian @ 1.062144
9 English_Kent_GBR + Icelandic + Irish + Macedonian @ 1.067125
10 British + British + Dutch + Montenegrian @ 1.073732
11 British + Gagauz + Norwegian_West + Scottish_Argyll_Bute_GBR @ 1.076506
12 Dutch + English_Kent_GBR + Montenegrian + Scottish_Argyll_Bute_GBR @ 1.082168
13 Dutch + Icelandic + Kosovar + Norwegian_West @ 1.096214
14 Icelandic + Irish + Kosovar + Swede @ 1.099154
15 Belgian + Frisian + Montenegrian + Scottish_Argyll_Bute_GBR @ 1.100481
16 CEU + Gagauz + Norwegian_West + Scottish_Argyll_Bute_GBR @ 1.100523
17 British + Icelandic + Irish + Macedonian @ 1.100630
18 Kosovar + North_German + Norwegian_West + Scottish_Argyll_Bute_GBR @ 1.100802
19 British + Dutch + Macedonian + Orcadian @ 1.105479
20 British + CEU + Dutch + Montenegrian @ 1.109074

Salkin
09-23-2014, 07:29 AM
Here's an example using MDLP K23b. My previous estimate of .05% difference is obviously off for this particular tool where it ranges from .1 to 1.12% difference; but I think you would agree there's not a whole lot of difference in the ancestral population percentages.

Indeed, doesn't look bad at all. Thanks for clearing that up.

I probably just latched on to a statement made somewhere about one particular calculator (and possibly just in one, early stage of its development) and overgeneralised it to everything GEDmatch and Oracle.

ArmandoR1b
09-23-2014, 12:25 PM
All my presumed positives in ftdna tree are all negative SNP's for me

That is not true. We have discussed this at length on two different forums. Some of the presumed positive in the screenshot at http://www.anthrogenica.com/showthread.php?3083-NatGeno2-questions&p=50393&viewfull=1#post50393 were not tested by Geno 2.0. That does not mean a negative SNP.

You still have a deep misunderstanding of the issue causing you to spread misinformation. Please stop.

ArmandoR1b
09-23-2014, 12:33 PM
For all of you that show Iberian ancestry in AncestryDNA it is because southern Europeans and northern Europeans share ancestry from both the Mesolithic and Neolithic migrations. Iberians also show to have a significant amount of north European ancestry at AncestryDNA even though there they have almost no north European ancestors in going back to the 16th century.

If you run your raw DNA files through Eurogenes K36 you will also see Iberian in your results. As stated in different words above, it is not because of a recent Iberian or southern European ancestor.

Erik
09-23-2014, 02:06 PM
For all of you that show Iberian ancestry in AncestryDNA it is because southern Europeans and northern Europeans share ancestry from both the Mesolithic and Neolithic migrations. Iberians also show to have a significant amount of north European ancestry at AncestryDNA even though there they have almost no north European ancestors in going back to the 16th century.

If you run your raw DNA files through Eurogenes K36 you will also see Iberian in your results. As stated in different words above, it is not because of a recent Iberian or southern European ancestor.
The calculator effect or actual Iberian ancestry is another possibility for me.

Scarlet Ibis
09-23-2014, 09:26 PM
Speaking of GEDmatch and blogger's ethnicity calculators, which do you recommend the most? I haven't had too much success with them.

Honestly, I'm not really sure about recommendations, because I think a holistic approach, where you take all plausible results into consideration, is generally the best. Personally, the ones that have seemed the most accurate for me have been the newer Eurogenes ones, and the new MDLP k23b. They're not perfect (nothing is), but commercial testing results have been more basic, and way behind the curve of genetics bloggers for years now. Good luck with your discoveries, anyway. This stuff can become addicting for years. ;)

Erik
10-11-2014, 04:05 PM
I just got an AncestryDNA kit for my first cousin on my mother's side. His dad is a mix of Irish, British, German and Norwegian.

dp
10-14-2014, 05:59 PM
The calculator effect or actual Iberian ancestry is another possibility for me.

[redacted per request]
Dear Erik,
here's a link ( http://dx.doi.org/10.1101/009340) that I think will interest you. The conclusions were based on 23andme kits.

ArmandoR1b
10-30-2014, 01:42 AM
The calculator effect or actual Iberian ancestry is another possibility for me.
[REDACTED]


What I described as the problem is partly the cause of "the calculator effect". Because some people inherit certain autosomal SNPS and other don't and the ones that didn't are included in the calculator then the ones that did are assigned a different ancestry. That is what the calculator effect basically is. Either your inherited DNA of your known ancestry is included with the reference population and doesn't get the calculator effect or it isn't and you get the calculator effect. You inherited DNA that was assigned to Iberians in the calculator instead of populations that you know you come from. The opposite actually happens with Spaniards and Latin Americans. They get north European results when they don't have known ancestry from that region in the past 500 years. This is all because of the mixing of Europeans over thousands of years and the AIMs (Ancestry-informative markers) go back 20,000 years or more. It takes many thousands of years of isolation for autosomal DNA to experience enough drift for AIMs to be correctly identified without any chance of assignment to a different ethnic population. The Iberian in the Ancestry test and Southern European in the FTDNA result could still be real but what I have explained could also be the cause like it is for many other people.

dp
10-31-2014, 02:52 PM
...
Glad to see all went well with their tests. I wonder how the GEDmatch calculators will treat her Finn 13% component.
dp