DNAprint accurate test?

[CaptDNA]

I took my DNAprint about 1 year ago. My results; 78% European 22% Native American.
Later I took a sample from my father. The results were a big surprise to both of us. His results: 56% European 28% Sub-Saharan African and 16% Native American!

I think we have a problem here. My Y-DNA test is a match with my fathers Y-DNA, so I feel better on that area ;-)
I don't think the results are accurate if my father DNAprint is 28% Sub-Saharan African. Where is my African percentage? I double check my fathers ancestors one by one (7 generations) and can't find the Sub-Saharan African ancestor. The only explanation to me is that one of our test is wrong. My father and I, also share the same haplogroup J.

Do we have a problem or it is possible?

Thank you

[JohnSanford]

Re: DNAprint accurate test?

[QUOTE]Originally posted by CaptDNA
[B]I took my DNAprint about 1 year ago. My results; 78% European 22% Native American.
Later I took a sample from my father. The results were a big surprise to both of us. His results: 56% European 28% Sub-Saharan African and 16% Native American!
. The only explanation to me is that one of our test is wrong.

Hello CaptDNA,
You appear to be right, "one of our test is wrong". They cannot both be correct. Do you realized that in fact, THEY BOTH MAY BE WRONG ? And where does that leave you, and how much have you spent?
-

[CaptDNA]

Re: Re: DNAprint accurate test?

[QUOTE]Originally posted by JohnSanford
[B]

Originally posted by CaptDNA
I took my DNAprint about 1 year ago. My results; 78% European 22% Native American.
Later I took a sample from my father. The results were a big surprise to both of us. His results: 56% European 28% Sub-Saharan African and 16% Native American!
. The only explanation to me is that one of our test is wrong.

Hello CaptDNA,
You appear to be right, "one of our test is wrong". They cannot both be correct. Do you realized that in fact, THEY BOTH MAY BE WRONG ? And where does that leave you, and how much have you spent?
-

Hello John

Perhaps this is a rare ocurrance, don't know.
The test is done by the same company Ancestry by DNA.
Quality Control vs Contamination. Possible?
I undestand that not all the descendants (brothers or sisters) will share the same percentages on the DNAprint, but this percentages are to significan to be a correct. The only thing I can do on my side is review my documentation for possible omissions or mistakes.

[AngelaC]

Re: Re: Re: DNAprint accurate test?

Originally posted by CaptDNA
Hello John

Perhaps this is a rare ocurrance, don't know.
The test is done by the same company Ancestry by DNA.
Quality Control vs Contamination. Possible?
I undestand that not all the descendants (brothers or sisters) will share the same percentages on the DNAprint, but this percentages are to significan to be a correct. The only thing I can do on my side is review my documentation for possible omissions or mistakes.

Hi,
I think that we cannot say for sure that the results are not genuine.

There are two issues here, - firstly, how does your father have african dna (& no sign of it in ancestry ), and secondly if your father does have subsaharan genes, how come you don't?

For the first issue, - I am not sure. Autosomal genes can hang around for a while. It would not be impossible to have genes from an ancestor generations ago (just less probable). It is more probable that the ancestor was just a few generations ago, but not impossible that was from back further.


For the second issue - The first thing that can be done to resolve this is to have a DNA print done on a sibling (if possible).
Hopefully, I'm not about to confuse you too much with my explanation......

We receive 50 % of our autosomal dna from each biological parent, thats a given.
When we go down further generations, this is where things become less predictable.
Mathematically, the probabilities are 25% of our autosomal DNA from each grandparent, 12.5% from each grandparent etc.
Those are the most probable results.

In reality, when the chromosome numbers are being split in half, it is totally random which grandparents chromosome you get (as well as that, there is gene swapping between each chromosome pair).
In reality you might have 10% from one grandparent,40% from one grandparent, or 50% from one grandparent & 0% from another.

With enough "samples" (ie. siblings tested) It all averages out to 25% from each grandparent, so if you tested all your siblings most would have about 14% subsaharan dna, some with 0%, some with 28%.


Angela.

[JohnSanford]

Re: DNAprint accurate test?

Originally posted by AngelaCP
Hi,
I think that we cannot say for sure that the results are not genuine.

There are two issues here, - firstly, how does your father have african dna (& no sign of it in ancestry ), and secondly if your father does have subsaharan genes, how come you don't?
Angela.

Hi AngelaCp, and CaptDNA,
And, I know we cannot say for sure, that the results ARE genuine! CaptDNA thinks that one or the other of them is WRONG, as the result of his and his Father’s DNAprint results – and the explanation of them. I say that one is wrong for sure, and there is a reasonable chance that BOTH are wrong. The testing laboratory could and should run the tests again (retest) both of them without charge, to indicate what in their opinion is the right answers, to remove some of this doubt.
I am amazed at the lengths that some will go to, to rationalize the value of tests of this kind. Much of it appears to be ‘wishful thinking’, if not unsupported rationalizations, from sincere believers. Sincerely, John Sanford

[AngelaC]

Re: Re: DNAprint accurate test?

Originally posted by JohnSanford
Hi AngelaCp, and CaptDNA,
I am amazed at the lengths that some will go to, to rationalize the value of tests of this kind. Much of it appears to be ‘wishful thinking’, if not unsupported rationalizations, from sincere believers. Sincerely, John Sanford

John, Sorry, you misunderstand me.
my "unsupported rationalisations" are not those of a "Sincere believer" who cannot be convinced that something is wrong. I am basically stating scientific facts, which I learned through 10 years of scientific training. Believe me, my levels of scientific scrutiny are very high, and if I thought that these tests were invalid, I would say so. Scientific results are never 100% black and white issues (more like 95 or 99%).

What I was trying to say to Captn DNA is that it is not impossible for those two results to be both correct, - it is unlikely but not impossible, based on the scientific facts on how genetic recombination occurs. In his case where his father shows 28% african, the most probable result for Captn DNA is 14%. It would be very rare to get a result of 0% but not 100% impossible. As a comparison - Winning the lottery is highly improbable, but not impossible (since somone almost always does win).

The results of one or both DNAprint tests might be wrong, but then again it is also possible that captn DNA has "won the DNA print lottery" so to speak. Testing a sibling of Captn DNA would resolve this issue (as it would be VERY improbable for both siblings to get 0% if the father is genuinely 28%)
Respectfully,
Angela.

[JohnSanford]

Re: Re: Re: DNAprint accurate test?

Originally posted by AngelaCP
Believe me, my levels of scientific scrutiny are very high, and if I thought that these tests were invalid, I would say so.
What I was trying to say to Captn DNA is that it is not impossible for those two results to be both correct, - it is unlikely but not impossible . . . . . . The results of one or both DNAprint tests might be wrong, but then again . . . . . Testing a sibling of Captn DNA would resolve this issue Angela. [/B]

Angela, Do you really think and believe that, "another test would RESOLVE this", or anything ??
On the subject of the “Merit” of DNA tests and ‘Testing results”, I have come to a conclusion: First I will say that I have seriously researched my own family genealogy for more than 25 years. I have a data-base that has over 16,140 different Individuals and over 38,500 documented Events that are attached to them. These persons are ALL related in some way to our children (Being the ancestry of both myself and my wife, of 57 years.). In all of this time I cannot recall ever adding any information that ever came from, or was reported to have been determined by, a DNA test of any kind.
How can you be so sure that the DNA tests are really useful for Genealogy purposes, when I have been advised that they cannot even absolutely assure us, through DNA testing, that even 4 living brothers of two true and faithful parents, will all come up with a 25/25 report in a ‘25 marker’ test? It is interesting, exciting, and mind boggling, BUT, is it USEFUL to our serious Family History searches, and the records that we develop for our children and their children?
It is my opinion that THEY ARE NOT. John Sanford, of CA.
-

[AngelaC]

John, CaptnDNA

Yes I really honestly do believe that a test with a sibling will resolve this, and my reasoning is scientific.

Firstly we should remember that these DNA prints are based on autosomal DNA ( the DNA that gets mixed up in every generation). So this type of test is very different to the Y-chromosomal and mtDNA tests (which are not mixed up down the generations).

Why do I honestly think that another test will resolve this issue?
I'll go through my logic....(according to mathematical principles of probability - which do apply to genetic recombination),

CaptnDNA fathers result says 28% subsaharan
Hypothetically, if Captn DNA's father had 100's of children, the vast majority would have half that subsaharan DNA (14%). some would have less, some would have more. There would also be a very very small number with none of that subsaharan DNA, and a very small number with all of that subsarahan DNA (ie. their result would show 28%, like their father). The chance of being one of those people with 0% of the subsararan DNA would be low, but like winning the lottery, it is not impossible.

Assuming that the fathers 28% result is genuine, - If a sibling of Captn DNA was tested, the probability of that sibling also getting a 0% would be incredibly low, - so if a sibling also got a 0% reading, we could comfortably conclude that the fathers result is wrong.

John,
As for your conclusions on the merit of these tests. Firstly, I will state a fact - geneology DNA tests have only been available for the last 4 years. The science behind them was originally developed for answering questions about ancient anthropological migrations. Since then, testing has become available to the public because the public requested that this type of test be made available. Yes, it is true that with a 25 marker test you cannot be 100% sure that a father and son will match, - however these small father to son DNA changes are the cornerstone of these tests. Without these very small changes, there wouldn't be any differences to detect.
True this lack of 100% precision can makes things tricky to interpret - .however, if two men have completely different results we CAN 100% for sure say they are not related. Proving people are related within a set number of generations can be tricky, but proving someone is totally unrelated is very easy.
For some names/family trees, this DNA testing does definely have merit for resolving questions. JABlair and myself have provided you with examples of some of the kinds of questions we are trying to answer with DNA testing. For other names/family trees, there may be no mysteries, nor issues to resolve. In those instances, geneology testing may be of no merit for that specific name/family tree - however, it does not negate the value of these tests for other names/family trees.
From what you say, you have a well documented family tree, with no "mysteries" you specifically would like to have solved. The only merits in DNA testing for you, I suspect, would be detecting relationships outside your well documented family tree. For instance, Are you interested in the ancient anthropological origin of your Y-chromosome or mtDNA??? Do all Sandfords in the world have a common origin?,
It might be that in your specific case, there is no merit in DNA testing for your family tree. Thats fine, - not all family trees are helped by DNA testing. However, other names/family trees are.
Angela

[JohnSanford]

Originally posted by AngelaCP
John, CaptnDNA

Yes I really honestly do believe that a test with a sibling will resolve this, and my reasoning is scientific. Angela

- - -
It is a fact that most of the people in the world actually believed passionately, that the earth was flat. Even those with the highest intelligence, and with the most education. But they were wrong, and most of them never learned or accepted it, Historians tell me that many still believed it long after it was proven that the earth is in fact ’round’. - So much for passionate belief that is sometimes faulty.
You asked the question "Do all Sandfords in the world have a common origin?" Our name is SANFORD, and I have documented records to prove it for over 18 generations. This is a good example of how false assumptions can affect advice , even from those who are highly intelligent and well informed. And, by the way, why would it be important to know if all people with any surname, be all related???
I believe that you are right when you say, all will not be helped by DNA searches - - that is exactly what I have contended. With kind regards, John Sanford, of CA.
- - -

[AngelaC]

Originally posted by JohnSanford
- - -
It is a fact that most of the people in the world actually believed passionately, that the earth was flat. Even those with the highest intelligence, and with the most education. But they were wrong, and most of them never learned or accepted it, Historians tell me that many still believed it long after it was proven that the earth is in fact ’round’. - So much for passionate belief that is sometimes faulty.
You asked the question "Do all Sandfords in the world have a common origin?" Our name is SANFORD, and I have documented records to prove it for over 18 generations. This is a good example of how false assumptions can affect advice , even from those who are highly intelligent and well informed. And, by the way, why would it be important to know if all people with any surname, be all related???
I believe that you are right when you say, all will not be helped by DNA searches - - that is exactly what I have contended. With kind regards, John Sanford, of CA.
- - -

Sorry John ,
I have a friend with the surname Sandford, I apologise for my error. My brain read Sanford as Sandford. I again apologise.
If all Sanfords in the world are descended from a comon ancestor 18 generations ago, you are very lucky to have a surname with only one origin. Many other people have surnames with multiple founders (like Smith etc.), and other surnames have changed down the ages (for instance some "Cones" are now Coans).
As you say, not all will be helped by DNA testing, - but others (eg. the many users of this service, including myself) are helped.
Regards,
Angela Cone-Phipps.

[Don Potter]

Enough is enough

John,
I respect your opinion as we are all allowed to have one, but I for one am more than a little tired of your constant negativity. This message board is provide by FTDNA and they do genetic testing. If you don't believe these tests and think it might turn out to be like, as you said, that the world is really not flat, then why don't you spend your time writing on another message board that you do believe is providing "true" information?
You have mentioned your well documented family line several times and I am glad for you. However, more than one person on this board has been kind enough to provide you with different references to explain all one could desire to know about this subject (DNA testing/results). These DNA tests may not help you but you have also been, once again, provide (by different people) numerous examples/senarios that these tests would be beneficial for. I don't intent to respond to any other message from you and would advise others to do the same.
Don Potter

For all others: I am reminded of a certain phrase here "You can lead a horse to water, but you can't make him drink".

[JohnSanford]

Genealogy entries of CaptDNA, in data-base?

Angela, I am impressed by the knowledge and enthusiasm that you seem to have for DNA testing, and it’s value for genealogy searches and searchers like CaptDNA and myself. My question for you is:
: If you were CaptDNA, and you had seriously studied the information that you were given, as a result of this particular DNA Testing. Just what entries would you put in your well documented ‘data-base’ about yourself, and also the entries you would make for your father? Also, what would you enter in the area for “Sources and Documentation? I would like to see what positive additions can and would be made to your Family History record. Sincerely, John Sanford, of CA.


[Quote] Originally posted by AngelaCP. . . John, CaptnDNA. Yes I really honestly do believe that a test with a sibling will resolve this, and my reasoning is scientific. .Assuming that the fathers 28% result is genuine, - If a sibling of Captn DNA was tested, . . .- so if a sibling also got a 0% reading, we could comfortably conclude that the fathers result is wrong. . . . . - not all family trees are helped by DNA testing. However, other names/family trees are. Angela ]

[Jim Barrett]

Re: Genealogy entries of CaptDNA, in data-base?

Originally posted by JohnSanford
[B]Angela, My question for you is:
: If you were CaptDNA, and you had seriously studied the information that you were given, as a result of this particular DNA Testing. Just what entries would you put in your well documented ‘data-base’ about yourself, and also the entries you would make for your father? Also, what would you enter in the area for “Sources and Documentation? I would like to see what positive additions can and would be made to your Family History record. Sincerely, John Sanford, of CA.

John, I'm not Angela or CaptDNA, but to me your question is very easy. I'll record my DNA results right along with all of my other documentation. It is one piece of evidence. It will be recorded with Bible, birth, death, marriage, obits, and census information.

All of these records have been recorded by human beings. All of them are subject to error. An each will be treated as one more piece of evidence. My birth certificate contains an error. The official copy was changed from the correct information to false information. My grandfather’s death certificate is full of errors, even though the information was provided by a son who was a MD. I have Bible records that I know had information added generations after the event. And who hasn’t seen errors in obits and census records. There is an additional source of information I haven’t mentioned. It is our memory. I find errors in mine.

In my opinion, anyone who is serious about genealogy is barking up the wrong tree when they take any one piece of evidence and allow it to over rule all other evidence.

I trust FTDNA and I trust my DNA results, but it would not be accepted over all other information if it didn't agree with the other information.

[AngelaC]

Re: Genealogy entries of CaptDNA, in data-base?

Originally posted by JohnSanford
: If you were CaptDNA, and you had seriously studied the information that you were given, as a result of this particular DNA Testing. Just what entries would you put in your well documented ‘data-base’ about yourself, and also the entries you would make for your father? Also, what would you enter in the area for “Sources and Documentation? I would like to see what positive additions can and would be made to your Family History record. Sincerely, John Sanford, of CA.

Hi John,
First, I should point out that the DNAprint uses a different type of DNA to the Y-markers test and the mtDNA tests, and is useful for only for giving a rough estimate of ones anthropological/ "racial" ancestry . It uses the autosomal (non sex-chromosome) DNA that is mixed up down each generation. The DNA print results will never pinpoint specific ancestors, so you cannot used the "DNA print" test to join specific people in a family tree. The Y-chromosome marker tests on the otherhand can (and the mtDNA tests to a certain extent). The different tests are all different tools to answering slightly different questions about ones ancestry - its not just about connecting people on the tree.

At this stage, Captn DNA cannot yet make a firm conclusion about whether both tests are correct, or one is incorrect. When I pointed out that it is possible to get the sort of result he has, I wasn't saying outright that I beleive 100% that both tests are correct, just that it is scientifically possible to get that result. Before Captn DNA can make firm conclusions about his fathers 28% Subsaharan result, then he does need to verify that the result is correct. There are two things he can do, request that both samples are retested (but if the results come back the same as originally done, you do have to pay something), and/or get a sibling tested. Untill something like that is done Captn DNA cannot yet make any conclusions about the result. If after testing a sibling, that test shows 0%, captn DNA then ignores the subsaharan result (no need to record anything in family tree), if the sibling shows from 10-28% subsaharan dna, then Captn DNA can conclude that there is subsaharan ancestry. Ok, ...how does he record this in his family tree then???......

First, he wouldn't use the test for connecting people together in the family tree. Many people present their family trees in book form, with appropriate text written about the family history, - origins etc, family stories. If they add the information to their family tree from DNA print tests it would be to state that, for instance, "there was a family legend that ancestor X might have had a Native American origin, and DNA print tests now suggest that this person might have had Subsaharan ancestry as well". The source information for this statement would be included.

As a personal example, I have an ancestor "Mary Anne" who was, according to family legend, Spanish. Documents (her fathers birth certificate) now show that her paternal grandfather was a Sephardic Jew, and her paternal grandmother was "a free mestee woman". Officially "mestee" was defined as 1/8 negro, but could also be sometimes used informally for different admixture levels. That means she might have been part Native American as well. I will always have a lingering wish that I knew with greater clarity information on this womans background...I will always wonder if I have a little bit of Native American ancestry. If it were phyically possible to get a DNA print test done on my ancestor "Mary Anne", then this question will be answered. It wouldn't change anything about who is recorded on my family tree or the family relationships, - but that information could then be included as part of the recorded family history text.

DNA print tests won't "connect the dots" on the family tree, but add to the recorded family history, and to a persons understanding of "where their ancestors came from".


Angela.

[CaptDNA]

Originally posted by AngelaCP
Sorry John ,
I have a friend with the surname Sandford, I apologise for my error. My brain read Sanford as Sandford. I again apologise.
If all Sanfords in the world are descended from a comon ancestor 18 generations ago, you are very lucky to have a surname with only one origin. Many other people have surnames with multiple founders (like Smith etc.), and other surnames have changed down the ages (for instance some "Cones" are now Coans).
As you say, not all will be helped by DNA testing, - but others (eg. the many users of this service, including myself) are helped.
Regards,
Angela Cone-Phipps.

Hello John and Angela

Thanks for helping with my dilema. I appreaciate both points of view and possibilities for my unusual and un expected results. I like to share with both an extract of a letter I got from AncestrybyDNA related to my questions and results.

"The reason we developed a new test with more markers is because we observed 6 cases (out of the 3,500 or so we have handled) that were troubling. For these 6 cases, we feel that the confidence space reported by the test on the triangle plot was reasonable, but not entirely correct, but more of a problem is that the MLE selected by the program using 71 markers was clearly not the best choice for these individuals (who had extensive genealogical histories documented, and results that differed among their family members). To make a long story short, the problem is not an algorithm problem, nor a problem with the markers we are using - but simply, it is a problem for rare individuals of high "heterozygosity" that is caused by not using enough markers for such people. To put it simply, our 71 marker test has a tough time with individuals of exceedingly high heterozygosity. Both you and either you or your mother (gbg*****) have high heterozygosity - in fact your level (52%) is the third highest value we have ever seen, and your moms is also in the top 10! High heterozygosity is something that is seen for certain individuals of rich admixture - for example a person that is 1/3rd EAS, 1/3rd EUR and 1/3rd AFR would have a very high heterozygosity, but such people are not all that common"

Once againg thanks John and Angela for helping undestand the results of DNAprint. Will see what my new test shows

Cheers