DNAPrint vs. decodeme vs. 23andme

[Starr173]

Has anyone been tested by all 3 companies or at least two? How do your results compare? I saw a post on the DNATribes log who scored 100% European on 23andme and 81% European, 13% African and 6% East Asian on decodeme. He is half Kabylie and 100% French on his mother's side. Please share your experiences with all 3 testing companies.

[rainbow]

This is a good topic. I wish there were replies. Sorry I can't be of any help.
I tested with only one of those companies, DNAPrint (AncestryByDna). I had ordered their Euro (1.0?) in 2006, but was tested for the 2.5 and was refunded the difference because I didn't qualify for the Euro. My result is 83% European and 17% Native American. I expected 100% European. I tested because I had wondered if I had any detectable Mongolian/Hun ancestry from my Slavic grandparent. If so, then I expected no more than 3% East Asian and 97% European. To this day the 17% is still a mystery.

[DKF]

This is a good topic. I wish there were replies. Sorry I can't be of any help.
I tested with only one of those companies, DNAPrint (AncestryByDna). I had ordered their Euro (1.0?) in 2006, but was tested for the 2.5 and was refunded the difference because I didn't qualify for the Euro. My result is 83% European and 17% Native American. I expected 100% European. I tested because I had wondered if I had any detectable Mongolian/Hun ancestry from my Slavic grandparent. If so, then I expected no more than 3% East Asian and 97% European. To this day the 17% is still a mystery.

At this point the only way you are going to know if your present results are valid or a false positive is to take the 23andme test. If you show up as any percentage whatsoever above 1% Asian, then you can rest assured that the DNAP test is NOT leading you down the garden path but is telling you something correct about your recent ancestry. They "set the bar" very high so everyone who is European tests as 100% European - as it should be. The decodeme test seems to be useful in obtaining information about ancient haploblocks which you share with other populations of the world.

[burto]

At this point the only way you are going to know if your present results are valid or a false positive is to take the 23andme test. If you show up as any percentage whatsoever above 1% Asian, then you can rest assured that the DNAP test is NOT leading you down the garden path but is telling you something correct about your recent ancestry. They "set the bar" very high so everyone who is European tests as 100% European - as it should be. The decodeme test seems to be useful in obtaining information about ancient haploblocks which you share with other populations of the world.

How does 23andMe differ to DeCodeMe in the sense of "predicting" ancestry?
Why do you say anything above 1% East Asian? Is that what the test picks up as being average for Europeans? Which do you think is better for finding out if you have more than one racial group in your ancestry and what about Native American?

[jloe]

How does 23andMe differ to DeCodeMe in the sense of "predicting" ancestry?
Why do you say anything above 1% East Asian? Is that what the test picks up as being average for Europeans? Which do you think is better for finding out if you have more than one racial group in your ancestry and what about Native American?

There's a little bit of a gray area here. It seems that many who test 100% European with 23andME test 90-95% European with decodeme.

I tested 90% European, 6% Asian, and 4% African with decodeme, while 100% European with AncestrybyDNA2.5.

The puzzle is that in some respects 23andME and decodeme are only using a subset of the 500,000+ and 1,000,000 SNPs resepctively that they test, approximately 20,000 SNPs for their ancestral determinations.

As David pointed out, decodeme tends to show more admixture than 23andME. In fairness, anyone 90% or more European on either test is phenotypically European and it is uncertain what the significance of that other ten percent really is. We await improvements in the analyses.

[Starr173]

Thanks for the replies. That's interesting what DFK said. That any admixture above 1% on 23andme probably reveals actual admixture since they set the bar so high that most Europeans score 100% Euro.

[juan carlos]

Thanks for the replies. That's interesting what DFK said. That any admixture above 1% on 23andme probably reveals actual admixture since they set the bar so high that most Europeans score 100% Euro.

That means that if 23andMe tells you you're 2% something, you can rest assured that is the case.

[Maria_W]

DNA Print vs DeCodeMe...

OK heres my results:

On DNA Print I recieved 10% Native American and 90% Indo European. (175 markers)
On DeCodeMe on my X chromosome I recieved 4% African American, 9% Asian and 87% Indo European.
On my Autosomal I recieved 2% African American, 6% Asian and 92% Indo European (1,000,000 markers)

Read carefully...Without including any of the above results from either test I have Potowomecke ancestry from the 1600's from Stafford Cty, VA. My mother (though her Waugh line) and it now looks like my father, who I had suspected last year before he died, found new information (though his Elkins line). They both come off of the 2 daughters of Wahangonoche, last weroance of the Potowomeke tribe. So it appears I have double lines of decent....

So what are the extra test results of decodeme saying?

Maria

[Starr173]

OK heres my results:

On DNA Print I recieved 10% Native American and 90% Indo European. (175 markers)
On DeCodeMe on my X chromosome I recieved 4% African American, 9% Asian and 87% Indo European.
On my Autosomal I recieved 2% African American, 6% Asian and 92% Indo European (1,000,000 markers)

Read carefully...Without including any of the above results from either test I have Potowomecke ancestry from the 1600's from Stafford Cty, VA. My mother (though her Waugh line) and it now looks like my father, who I had suspected last year before he died, found new information (though his Elkins line). They both come off of the 2 daughters of Wahangonoche, last weroance of the Potowomeke tribe. So it appears I have double lines of decent....

So what are the extra test results of decodeme saying?

Maria

I've seen your results for you and your family on these boards. Thanks for sharing all your results. It really gives others info on the different testing companies and how accurate they are at detecting admixture. From what I see, who is correct? 23andme detects the least admixture and finds Caucasians primarily 100% European.
I saw your Dad's results and I was wondering if he took the DNAPrint 2.5 or 2.0? I ask this because I've seen that the old 2.0 failed to detect admixture in many cases.
It seems like your X chromosome has more Native-American(Asian) than your autosomal. I guess since there are two X's you can't tell if its coming from your Mom or Dad.
But it's cool that an autosomal test can pick up admixture from the 1600's!

[DKF]

OK heres my results:

On DNA Print I recieved 10% Native American and 90% Indo European. (175 markers)
On DeCodeMe on my X chromosome I recieved 4% African American, 9% Asian and 87% Indo European.
On my Autosomal I recieved 2% African American, 6% Asian and 92% Indo European (1,000,000 markers)

Read carefully...Without including any of the above results from either test I have Potowomecke ancestry from the 1600's from Stafford Cty, VA. My mother (though her Waugh line) and it now looks like my father, who I had suspected last year before he died, found new information (though his Elkins line). They both come off of the 2 daughters of Wahangonoche, last weroance of the Potowomeke tribe. So it appears I have double lines of decent....

So what are the extra test results of decodeme saying?

Maria

Maria and all interested parties:

I have had a chance to examine the profiles of a number of people who have taken the decodeme, 23andme, and ABDNA testing (and that includes me).

Clearly if you "want" to find Native American ancestry one would be well advised to go to ABDNA for your testing. If someone is interested in the facts and not the romance, it will have to be 23andme testing.

In one instance, someone of European ancestry (but will a long Colonial USA pedigree) and no known Asian, Native American or African ancestry received a percentage of say 23% NA or Asian (I don't wish to identify the person). They insisted that he or she had an unknown grandparent who was not European. This has caused no end of speculation as to what the correct version is. The genealogy is crystal clear - no minority ancestry. But the DNA result was so clear and dramatic. Well, this person took the 23andme test and guess what.....drum roll please.........100% European.

Thus for those obtaining 10 to 20% Native American with any known NA ancestors back in the 1600s or undocumented then if it is the truth they seek, my suggestion is to take the 23andme test and see what a test with 20,000 ancestral informative markers will say. If they paint any part of any chromosome and even if only 1%, this would be considered likely valid. I have seen enough profiles of people with European heritage (always 100% European unless from Finland then 1% or less Asian is typical). For someone with a North Dakota ancestor from 8 generations ago, 2% Asian with markers sprinkled over a number of chromosomes. Others with lengthy Colonial ancestry have often come back with say 1% African or 1% Asian. Again, I would consider this significant (suggesting an ancestor who was African or NA within the past 200 or a bit more years).

I and I am sure many others would be pleased if those with little to no documented NA ancestry but with 10% or above on the ABDNA test would weigh in with their 23andme results - good news or not good news. For me only the truth will do. Even though I can see some haplotype blocks on the decodeme browser for African or Asian I believe my 100% European by 23andme. My NA ancestor born in 1747 was simply too far back for anything to show (although they have not painted the X chromosome yet where my largest blocks reside). I may have a relative who is 4 generations closer to the NA ancestor take the test and see what results emerge. It is always a bit of a risk if part of one's identity hinges on being part NA but in my opinion it does no one any good to put forward DNA results which essentially are wrong and misleading. Hopefully anyone in this category is up for the challenge - the results may confirm what ABDNA says, but who knows.

[Starr173]

Mr. Faux, what would you say of my 8% Native-American on DNAPrint 2.5? My ancestry is from Puerto Rico and the Native-American population died out in the early 1500's. By the 1600's they were no pure Native-Americans. So, I too, do not have any recent Native-American ancestry but DNAPrint was still able to pick up 8%.

[DKF]

Mr. Faux, what would you say of my 8% Native-American on DNAPrint 2.5? My ancestry is from Puerto Rico and the Native-American population died out in the early 1500's. By the 1600's they were no pure Native-Americans. So, I too, do not have any recent Native-American ancestry but DNAPrint was still able to pick up 8%.

Frankly, no one can tell you what results to expect. There could be sufficient haploblocks from NA ancestors if perchance in Puerto Rico you have an assortment of ancestors who were admixed, irrespective of the lack of anyone who was in the 100% NA range back to the 1600s. I would expect many Puerto Ricans to have a "sprinkling" of NA (and African) on their autosomes but I suppose one would see everything from 100% European to say 5% NA. The blessing of 23andme is that they even show any block that they consider Asian and it could be just a small blotch on one chromosome that is colored Orange. Bottom line is that it is a gamble and it is possible that even if you show as 100% some of your kin could have a small amount of NA - the luck of the draw and all that during the chromosome shuffling process.

There are so many additional features. For example for females mtDNA haplogroup, health and trait info, comparison to groups from around the world on half a million markers and placement shown by the whole genome. On the latter I fall most clearly within the NW German group but also on the border of the Norwegian, France, and England groups (only) - which absolutely fits knowing what I do about my ancestry. One can make direct comparisons with what has been published in the recent literature to see how this amazing technology works. For example, depending on N or S, Italians cluster in the Southern European group and likely within the Italian group, although the S Italians may find themselves overlapping in the Middle Eastern group which makes sense knowing the history of the area (and geography).

[Starr173]

Frankly, no one can tell you what results to expect. There could be sufficient haploblocks from NA ancestors if perchance in Puerto Rico you have an assortment of ancestors who were admixed, irrespective of the lack of anyone who was in the 100% NA range back to the 1600s. I would expect many Puerto Ricans to have a "sprinkling" of NA (and African) on their autosomes but I suppose one would see everything from 100% European to say 5% NA. The blessing of 23andme is that they even show any block that they consider Asian and it could be just a small blotch on one chromosome that is colored Orange. Bottom line is that it is a gamble and it is possible that even if you show as 100% some of your kin could have a small amount of NA - the luck of the draw and all that during the chromosome shuffling process.

There are so many additional features. For example for females mtDNA haplogroup, health and trait info, comparison to groups from around the world on half a million markers and placement shown by the whole genome. On the latter I fall most clearly within the NW German group but also on the border of the Norwegian, France, and England groups (only) - which absolutely fits knowing what I do about my ancestry. One can make direct comparisons with what has been published in the recent literature to see how this amazing technology works. For example, depending on N or S, Italians cluster in the Southern European group and likely within the Italian group, although the S Italians may find themselves overlapping in the Middle Eastern group which makes sense knowing the history of the area (and geography).

Mr. Faux, thank you for your reply. It was very scientific and well-thought out. Why doesn't 23andme offer just the ancestry painting and the global similarity option for those of us who already know our mtDNA and y-dna haplogroups? I'm also not too keen on knowing my risks for various disorders. Knowing your family history can tell you almost the same thing. No I'm not completely sold on 23andme for those reasons.

[Noaide]

Why doesn't 23andme offer just the ancestry painting and the global similarity option for those of us who already know our mtDNA and y-dna haplogroups? I'm also not too keen on knowing my risks for various disorders. Knowing your family history can tell you almost the same thing. No I'm not completely sold on 23andme for those reasons.

They would then have to redesign the chip not to include Y-DNA and mtDNA, that would mean more hazzle and more cost so they might as well just include them in the same batch.

[Noaide]

Mr. Faux, what would you say of my 8% Native-American on DNAPrint 2.5? My ancestry is from Puerto Rico and the Native-American population died out in the early 1500's. By the 1600's they were no pure Native-Americans. So, I too, do not have any recent Native-American ancestry but DNAPrint was still able to pick up 8%.

I also have 8% NA on the ABDNA test, however none of my ancestors have never made any footprint in the new world.