Disappointment

[JPHutchins]

As an adoptee, I am very anxious to determine the identity of my bio-father. A couple of months ago I discovered a 25/26 match (GD=0) on YSearch.com. The characteristics, demographics and time frames matched what I was looking for.

The person had tested with Sorenson Molecular Genealogy Foundation (SMGF) and refused to take any further DNA test. I located a son of the person and he agreed to take an FTDNA Y-DNA37 test.

After the test was taken, the manager of the first person's DNA account realized that the conversion from SMGF to FTDNA had been incorrect. She corrected the figures resulting in a 20/26 match with a GD=6.

Last week I received the results for the first 25 markers from FTDNA. The match is 21/25; Not Good. The son's markers exactly matched those of his father.

Once again I am at square zero. It is really too bad that near matches don't mean anything. In an FAQ, FTDNA states:

A 21/25 match is too far off to be considered related within the genealogical time frame.

It is vaguely possible that the rule for Probably Not Related applies. The only way to confirm to relationship is to test additional family lines and to find where the mutations took place. By testing additional family lines, you can find the person in between. This 'in betweener' is essential for you to find. Without him, it is highly unlikely that the match is genealogical.

Wouldn't it be nice if when you got near matches, you were able to at least determine the surname or something that would direct an adoptee like me in a correct direction. I am aware that perhaps there in a MRCA 10 or more generations back, but how would that be tracked.

The bottom line is that it is very disappointing, but I am about 1/350,000,000 closer to finding my bio-father. One less person to test, one more entry in the FTDNA database.

[thetick]

I'm sorry to hear your story. It also is a real shame that you could have a match that does not show as a match due to a bad/lack of conversion from a different company. Ysearch should really fix this by providing the conversion, but they really don't care.

[Frederator]

Not to give you false hope, but are you sure that this other fellah got the conversion right? Going from GD 0 to GD 6 merely by correcting a conversion error seems out of whack to me--I think there are only a couple of markers in the relevant Sorenson/FTDNA range that differ in lab protocol; too few to generate that big a difference.

Except DYS389-2. That one you have to subtract the FTDNA results for DYS389-1 to get the Sorenson equivalent. Do you think it's possible that your friend screwed up the conversion instructions somewhere and subtracted when he should have added?

Could you post the relevant YSearch ID #'s so we can take a look? This really just doesn't seem to make sense to me. Unless you're from some rare haplogroup, I think the value of DYS389-2 is almost always like 13 or so off between Sorenson and FTDNA.

I have no results worth talking about either--six years on now.

Best of luck anyhow.

As an adoptee, I am very anxious to determine the identity of my bio-father. A couple of months ago I discovered a 25/26 match (GD=0) on YSearch.com. The characteristics, demographics and time frames matched what I was looking for.

The person had tested with Sorenson Molecular Genealogy Foundation (SMGF) and refused to take any further DNA test. I located a son of the person and he agreed to take an FTDNA Y-DNA37 test.

After the test was taken, the manager of the first person's DNA account realized that the conversion from SMGF to FTDNA had been incorrect. She corrected the figures resulting in a 20/26 match with a GD=6.

Last week I received the results for the first 25 markers from FTDNA. The match is 21/25; Not Good. The son's markers exactly matched those of his father.

Once again I am at square zero. It is really too bad that near matches don't mean anything. In an FAQ, FTDNA states:

A 21/25 match is too far off to be considered related within the genealogical time frame.

It is vaguely possible that the rule for Probably Not Related applies. The only way to confirm to relationship is to test additional family lines and to find where the mutations took place. By testing additional family lines, you can find the person in between. This 'in betweener' is essential for you to find. Without him, it is highly unlikely that the match is genealogical.

Wouldn't it be nice if when you got near matches, you were able to at least determine the surname or something that would direct an adoptee like me in a correct direction. I am aware that perhaps there in a MRCA 10 or more generations back, but how would that be tracked.

The bottom line is that it is very disappointing, but I am about 1/350,000,000 closer to finding my bio-father. One less person to test, one more entry in the FTDNA database.

[JPHutchins]

I think that I may have written this in a way that might make for some confusion.

The son tested with FTDNA, while the father tested with SMGF and had his results converted to meet FTDNA standards (there were problems with the conversion). I tested over a year and a half ago with FTDNA.

The son's 25 marker results have come back from FTDNA and match exactly the father's recently converted results. That would indicate that the converted results were converted correctly on the third attempt.

Frederator, I will send you a private message with the account numbers, but would prefer to not make the other accounts public. My YSearch account number is a part of my signature.

My disappointment is that the Y-DNA results are basically an all-or-none situation where near matches do not provide any type of road map.

[twang]

Well at least you know you're related to a lot of interesting people.

http://en.wikipedia.org/wiki/Haplogroup_R1b_%28Y-DNA%29

[nathanm]

Not to give you false hope, but are you sure that this other fellah got the conversion right? Going from GD 0 to GD 6 merely by correcting a conversion error seems out of whack to me--I think there are only a couple of markers in the relevant Sorenson/FTDNA range that differ in lab protocol; too few to generate that big a difference.

Except DYS389-2. That one you have to subtract the FTDNA results for DYS389-1 to get the Sorenson equivalent. Do you think it's possible that your friend screwed up the conversion instructions somewhere and subtracted when he should have added?

Could you post the relevant YSearch ID #'s so we can take a look? This really just doesn't seem to make sense to me. Unless you're from some rare haplogroup, I think the value of DYS389-2 is almost always like 13 or so off between Sorenson and FTDNA.

I have no results worth talking about either--six years on now.

Best of luck anyhow.

It's not out of whack. Converting from SMGF to FTDNA means the following additions:
DYS441 +1
DYS442 +5
YGATAA10 +2
GATA H4.1 +1
And the difference in DYS389. See this table for reference.

[dwight]

Very sorry to hear, it must be a terrible let down, John. Hopefully better things to come before too long!

Dwight

[Frederator]

Um, DYS389-1 and DYS389-2, which : the 389 you quote, is the only affected marker falling within FTDNA's first 25 markers--hence, absent any lab errors, such a result would indeed appear to be out of whack with a GD differential of 6 in the first 25 markers as described.

DYS389-2 less DYS389-1 equals DYS389a reported by many labs exactly as I described.

It's not out of whack. Converting from SMGF to FTDNA means the following additions:
DYS441 +1
DYS442 +5
YGATAA10 +2
GATA H4.1 +1
And the difference in DYS389. See this table for reference.

[Ann Turner]

It's not out of whack. Converting from SMGF to FTDNA means the following additions:
DYS441 +1
DYS442 +5
YGATAA10 +2
GATA H4.1 +1
And the difference in DYS389. See this table for reference.

DYS389-2 is the same for SMGF and FTDNA (reporting the total length of DYS389). Oxford Ancestors and the Genographic project report DYS389ii-i (two minus one) or DYS389B as the length of the second part.

[Frederator]

Thanks, Ann--I could never remember which labs used that old Oxford protocol.

So it seems that NONE of the relevant markers appear within FTDNA's first 25.

Must be a lab error. Not to trash Sorenson, but I've heard of a number of lab errors over there--which is a shame, not only because it could result in nasty surprises like the one we may have seen here, but because I really like their DB interface. Still, a DB, no matter how fancy, can't mean much if your lab is throwing out 6 repeat errors willy-nilly....

DYS389-2 is the same for SMGF and FTDNA (reporting the total length of DYS389). Oxford Ancestors and the Genographic project report DYS389ii-i (two minus one) or DYS389B as the length of the second part.

[Ann Turner]

Thanks, Ann--I could never remember which labs used that old Oxford protocol.

So it seems that NONE of the relevant markers appear within FTDNA's first 25.

Must be a lab error. Not to trash Sorenson, but I've heard of a number of lab errors over there--which is a shame, not only because it could result in nasty surprises like the one we may have seen here, but because I really like their DB interface. Still, a DB, no matter how fancy, can't mean much if your lab is throwing out 6 repeat errors willy-nilly....

See message #4 in this thread. The DNA results from the two companies match if converted correctly. The original YSearch record (manually entered) was the source of the discrepancy.

I suspect what you describe as "lab errors" are mostly due to conversion issues. AFAIK, FTDNA still plans to adopt the NIST standard, but it's not a simple change to make because of ripple effects in various components of the database and website.

[Frederator]

Actually, no. The errors in this case are indeed lab errors. JP sent me the YSearch ID #'s, and none of the markers in question have any reporting differences related to them. Read my quote more carefully.

See message #4 in this thread. The DNA results from the two companies match if converted correctly. The original YSearch record (manually entered) was the source of the discrepancy.

I suspect what you describe as "lab errors" are mostly due to conversion issues. AFAIK, FTDNA still plans to adopt the NIST standard, but it's not a simple change to make because of ripple effects in various components of the database and website.

[JPHutchins]

The GD=6 determination was made in YSearch.com based on the markers tested with SMGF vs FTDNA. I have tested to a total of 74 markers.

Within the first 25 markers as tested by FTDNA, the GD for the father is 3 as SMGF does not test DYS 437. But SMGF tested 10 additional markers beyond the first 25 of FTDNA. Those 10 markers added an additional GD of 3 for a total of 6.

The son has only received the results from FTDNA 25 as of yet, with a GD=4. The difference is that the son's DYS 437 adds a GD of 1 over that of the father.

So when I originally wrote my posting I neglected to mention that there were additional markers beyond the FTDNA 25.

Sorry.

[Frederator]

But the upshot turned out to be the same when I looked at the profiles: None of the markers in question were associated with reporting protocol differences. Only one of the differences was outside the first 25 markers.

Assuming that the father's haplotype was entered into SMF's database by SMF (and from everything you've told me, there's no indication that his original test was done elsewhere, so I have to assume that SMF did), it looks pretty clear to me that the source of the problem is a lab error on SMF's part.

Whether the error was in the performance of the test, interpretation of the test or clerical recording of the results, I think it all adds up to lab error on the original SMF test.

And getting to the difference between GD returned by the Ysearch DB and the Infinite Allele model, it is just possible that a more accurate interpretation would be GD 5 rather than 6. From time to time a mutation event will result in a marker changing by more than one repeat. This could still carry some significance if you didn't turn up any additional differences on a full 67 marker test.

Mutations are random events, and despite the best efforts of the scientists who selected the particular markers for each panel to obtain the best-balanced, most-representative array of markers possible, it could be that your match quality improves at a higher resolution. That certainly is the case with all of my "near matches"--nothing hits the radar until I've tested 37 markers.

But whether you and your friend upgrade to 67 markers is entirely up to you. To be honest, my gut feeling is that it wouldn't help too much. I could be wrong, but I think the best case scenario is probably that you match like 8 out of 67--might mean a common ancestor around the mid-1500's or something, but even that would be far too loose a match to say so definitively.

And even if you don't go the full 67, I think that this test could have some vague significance for you. These databases are very Ireland- and Britain-heavy--there just aren't enough continental Europeans represented yet. Given the vastly larger population of continental Europe and this skewing of the database, even a GD of 6/5 at 37 does suggest that your remote ancestors originated in the same part of the world. I realize that's not the answer you want or need to get to, but it IS something to think about.


Best of luck.

The GD=6 determination was made in YSearch.com based on the markers tested with SMGF vs FTDNA. I have tested to a total of 74 markers.

Within the first 25 markers as tested by FTDNA, the GD for the father is 3 as SMGF does not test DYS 437. But SMGF tested 10 additional markers beyond the first 25 of FTDNA. Those 10 markers added an additional GD of 3 for a total of 6.

The son has only received the results from FTDNA 25 as of yet, with a GD=4. The difference is that the son's DYS 437 adds a GD of 1 over that of the father.

So when I originally wrote my posting I neglected to mention that there were additional markers beyond the FTDNA 25.

Sorry.

[DAStratton]

Frederator, I hope you are also using some of the adoptee search groups out there and not just depending on DNA. Making a one-on-one match like that with a database that still is limited basically to people who voluntarily test and register is like a needle in a haystack. Many of the volunteers at the adoptee search groups are as good as professionals, and help others because they have been there, on one side or the other. I have seen them do amazing work. I think there is an online group for every state and country, and if you just have the date and place of birth, even that can be enough for them to start with. DNA can then be used to validate a find, if needed. Send me a private message if you'd like.