PDA

View Full Version : Z18133 results in Geno 2.0 NG tests - possible false positives



swid
09-12-2016, 05:35 PM
Based on two examples in my R-ZZ10 project, I now have a strong suspicion that anyone who is listed as R-Z18133 (which is several levels below Z253 in reality) based *solely* on Geno 2.0 NG testing is likely to be an incorrect place on the haplotree.

N193040 - the next SNP up the haplotree that shows as tested positive is L21; an exact 37-marker match with the same surname and contact email (who is also in the project) has been confirmed as Z253- by both the R1b Backbone and L21 SNP Packs.

477741 - the next SNP up the haplotree that shows as tested positive is L51 (he has P310 in the raw results as well); unfortunately, this man has a fairly vanilla R1b haplotype and has only tested 25 markers, so I don't have any informed guesses from STR patterns as to what his actual SNP grouping might be.

I logged a ticket to FTDNA to have them look at those two results.

ArmandoR1b
09-12-2016, 07:59 PM
I'm not showing Z18133 to be a false positive in the NextGen results of a U106 or of a HG-I person. It doesn't surprise me though that there are people that are positive by chance for an individual SNP and are assigned that subclade even though their upstream SNPs show they are negative for them. It's still weird that you have two participants with false positives but I have two files that don't.

Have you downloaded their Y-DNA SNPs as CSV using the link as shown in the following screenshot?

http://i.imgur.com/gcRTGZr.jpg

Once you download the CSV file you can search for the results of Z18133 and Z253 and see if they are ancestral for both Z18133 and Z253 or just Z253. CTS1751 and L21 are also tested by Geno 2.0 NG.

swid
09-12-2016, 09:05 PM
Thanks! As I've never had any Geno 2.0 testing myself, I didn't realize that there was *another* raw data file to download immediately below the SNP download link in the Y-DNA section. I've downloaded the .csv files for both.

The file for 477741 is completely blank, unfortunately, so I don't have anything more to go on for him. However, the file for N193040 has SNP calls in it. Z18133 has the ancestral call of G there, as does Z253 (which is also G), so in that case the error seems to be fairly clearly in the FTDNA interpretation of the Geno 2.0 NG results.

lgmayka
09-13-2016, 02:53 AM
Simply looking at my project's SNP report (https://www.familytreedna.com/public/polish?iframe=ysnp), I see about 20 false Z18133+ results:
#B79612, I-CTS1846
#354451, I-Y3548
#N163095, I-Z187
#N153825, J-L816
#N212163, J-Z600
etc.

ArmandoR1b
09-13-2016, 09:23 PM
Simply looking at my project's SNP report (https://www.familytreedna.com/public/polish?iframe=ysnp), I see about 20 false Z18133+ results:
#B79612, I-CTS1846
#354451, I-Y3548
#N163095, I-Z187
#N153825, J-L816
#N212163, J-Z600
etc.

If you download their CSV files they probably have G in their raw data.

ArmandoR1b
09-13-2016, 09:34 PM
Thanks! As I've never had any Geno 2.0 testing myself, I didn't realize that there was *another* raw data file to download immediately below the SNP download link in the Y-DNA section. I've downloaded the .csv files for both.

The file for 477741 is completely blank, unfortunately, so I don't have anything more to go on for him. However, the file for N193040 has SNP calls in it. Z18133 has the ancestral call of G there, as does Z253 (which is also G), so in that case the error seems to be fairly clearly in the FTDNA interpretation of the Geno 2.0 NG results.

I looked at the SNP page of the projects the people whose files I had looked at, that that have G in their raw data files, and they also show to be Z18133+. The DF27 SNP page (https://www.familytreedna.com/public/R1b-DF27?iframe=ysnp) also shows a lot of people with Z18133+ and they are probably also Geno 2.0 NG customers.

Most of the time that I have seen false positives it has been in the raw data and not solely in the SNP report page. I have been complaining for a long time about FTDNA not getting the negative calls from Geno 2.0 and that causing an incorrect placement of people into a subclade they don't belong in. I wish Geno 2.0 and FTDNA would collaborate a lot more on getting the problems fixed. It seems as though they need a programmer dedicated to creating an algorithm that sorts the results correctly based on both positives and negatives and filters out the false positives based on them being found in results from many different haplogroups.

I wonder how many other SNPs that are negative in the Geno 2.0 NG raw data are shown as positives at FTDNA. Your project seems to be affected the most due to L21 being positive.

swid
09-16-2016, 07:09 PM
One step forward, one step back.

N193040 now displays as being R-S7123 (which is an SNP in the L193 block). The S7123 mutation is A-> G; he has a C call in the raw data. He also has a negative call for Z16332 and a C call for the G -> A mutation for S5992 (both of which are also in the block). Of course, he's also ancestral for other SNPs going upstream towards L513.

Geno 2.0 results are a mess. :deadhorse:

SteveJ
09-18-2016, 07:17 PM
N193040 is my father's kit (I'm N129953, ZZ10_1), and I was understandably a bit perplexed when I transferred his data over from NatGeo. It's good (figuratively speaking) to now know that this discrepancy isn't just an isolated incident, but a known issue with Geno 2.0 NG interpretation results that can possibly (hopefully) be remedied in the future. My plan is to continue testing his kit as finances allow, but I wonder if it's more feasible to wait and see if this issue gets corrected first.
Thoughts?

swid
09-18-2016, 10:25 PM
N193040 is my father's kit (I'm N129953, ZZ10_1), and I was understandably a bit perplexed when I transferred his data over from NatGeo. It's good (figuratively speaking) to now know that this discrepancy isn't just an isolated incident, but a known issue with Geno 2.0 NG interpretation results that can possibly (hopefully) be remedied in the future. My plan is to continue testing his kit as finances allow, but I wonder if it's more feasible to wait and see if this issue gets corrected first.
Thoughts?

Steve, I'm the admin of the R-ZZ10 project; I'm glad you showed up on here to clarify the relationship among those kits. Welcome! :)

Hopefully FTDNA will get their act together on this; it may help for you to log a ticket to their help desk as well. Since Geno 2.0 NG results don't include DF13 (I think), your father's results should display as R-L21 when they get everything sorted out.

In any case, since you already know you're ZZ10+ and Z253-, Z255-, CTS3386-, and MC14- from your own results, there's not a whole lot left to test for either one of you. For selfish reasons, I'm curious if you'd turn out to be a member of my own subclade -- BY4046. At this time, the only way to find that out would be to order a Big Y or Full Genomes test (either of which is pretty expensive, but would basically be the last test you'd ever need to order for either kit) or to order BY4047 (which is in the same block as BY4046) from YSEQ. If you've never tested there, the total cost of that would be $22.50.

Finally, if you simply want your father's FTDNA results to match up with yours, you can order R-ZZ10_1 as a standalone SNP test for $39.

rivergirl
01-21-2017, 01:07 AM
I have a member who has been tested with Geno 2, and was CTS11751+. He matches a CTS1751 cluster so have no reason to think he isn't.
His kit was batched to the lab this week for Big Y testing. Today he shows as R-S7123.
I take it that this is some sort of glitch with FTDNA system.

His BIg Y test is still showing awaiting results.

wombatofthenorth
01-21-2017, 05:27 AM
For some reason FTDNA doesn't seem to know when Geno uses reverse strand all the time and so on. A few mistakes can appear on the Geno site itself I think, but it seems like a great many of the problems don't appear on the Geno site but only with how FTDNA tries to handle transfers. Not sure why FTDNA seems to use a different translation book than Geno does. That is especially odd since FTDNA made the chip and carries out the actual tests for them....

ArmandoR1b
01-21-2017, 02:23 PM
I have a member who has been tested with Geno 2, and was CTS11751+. He matches a CTS1751 cluster so have no reason to think he isn't.
You could download the CSV file as pointed out in the example in post #2 (http://www.anthrogenica.com/showthread.php?8553-Z18133-results-in-Geno-2-0-NG-tests-possible-false-positives&p=186306&viewfull=1#post186306) then open it and then look up the following SNPs that are phylogenetically equivalent to CTS11751 and do not show up as false positives in Geno 2.0 raw data. If the SNPs below have the allele on the left then he is negative. If he has the allele on the then right he is positive.

SNP Ancestral (Negative) Derived (Positive)
L80 A G
L157 T A
M253 C T
L187 A T
P30 G A
L845 T G

What is his result for those SNPs?

Does he have CTS11751 in his results? I can't find that SNP in the kits that I have.

The following is a false positive in Geno 2.0 raw data for people that do not belong to the I1 haplogroup.

SNP Ancestral (Negative) Derived (Positive)
L840 C G


His kit was batched to the lab this week for Big Y testing. Today he shows as R-S7123.
I take it that this is some sort of glitch with FTDNA system.
The negative results from Geno 2.0 are not transferred to FTDNA. This has been a complaint of mine for a long time. If the negative results were transferred to FTDNA then S7123 could be ignored regardless of what the result for it is. Anyone can have mutations found in other haplogroups but if the upstream SNPs are negative then the mutation has no meaning.

S7123 is a mutation of A to G but Geno 2.0 raw data has a C for the kits that I have the results for. Therefore the result is useless regardless of the reason that a C is showing up.

These are some of the SNPs upstream from R-S7123 that he would be positive for if he really belonged to the S7123 subclade.

SNP Ancestral (Negative) Derived (Positive)
P310 A C
L52 C T
L51 G A
L21 C G
CTS5396 T C
Z249 T C
S5668 A G
FGC13499 G A
Z16332 A G
A6 C T

What are his results for those?


His BIg Y test is still showing awaiting results.
You should have the BigY results in less than a month.

ArmandoR1b
01-21-2017, 02:28 PM
For some reason FTDNA doesn't seem to know when Geno uses reverse strand all the time and so on.
Geno should not be using results from the reverse strand since none of the Y-DNA databases such as Ybrowse and ISOGG use the reverse strand and none of the other DNA companies such as FTDNA, 23andme, Yseq, and BritainsDNA use the reverse strand as a valid result. They are useless results no matter what.

wombatofthenorth
01-21-2017, 10:36 PM
Geno should not be using results from the reverse strand since none of the Y-DNA databases such as Ybrowse and ISOGG use the reverse strand and none of the other DNA companies such as FTDNA, 23andme, Yseq, and BritainsDNA use the reverse strand as a valid result. They are useless results no matter what.

they are not useless if you know what is what

anyway it's FTDNA he gives the raw data out to the public, they designed the chip for them so it seems odd they don't know what they did for what snp even when Geno has a better idea somehow, why does FTDNA not give uniform strand results?

rivergirl
01-22-2017, 02:20 AM
Thanks ArmandoR1b, the SNP should have read CTS1751, (my typo put CTS11751) CTS1751 is under R-L21, DF13, Z38589

Here are his results;

SNP Ancestral (Negative) Derived (Positive)
L80 A G -- not found
L157 T A -- TT
M253 C T -- CC
L187 A T -- AA
P30 G A -- not found
L845 T G -- TT

L840 GG -- GG

SNP Ancestral (Negative) Derived (Positive)
P310 A C -- CC
L52 C T -- TT
L51 G A -- AA
L21 C G -- GG
CTS5396 T C -- TT
Z249 T C -- TT
S5668 A G -- AA
FGC13499 G A -- CC
Z16332 A G -- AA
A6 C T -- not found

His CTS1751 is CC. I thought it should be A-G
S7123 is CC

I will wait for his Big y results. This is the 2nd time FTDNA have changed his haplogroup.

ArmandoR1b
01-22-2017, 01:26 PM
they are not useless if you know what is what
Supposedly only Geno 2.0 knows which are using the reverse strand and I have an extreme doubt they really are using the reverse strand for a lot of the problematic SNPs. Since none of the databases use the reverse strand it's someone's word. That's not good enough. On top of all of that, there are way too many false positives for all of them to be using the reverse strand.

Geno 2.0 NG did not use quality control like the Geno 2.0 and BritainsDNA Chromo2 did and LivingDNA is doing now. So any allele that does not match the ancestral or derived allele on the forward strand in ISOGG or Ybrowse should not be included in the results.


anyway it's FTDNA he gives the raw data out to the public, they designed the chip for them so it seems odd they don't know what they did for what snp even when Geno has a better idea somehow, why does FTDNA not give uniform strand results?
Geno should be the one giving out the raw data and not FTDNA and I'll say it again - there are way too many false positives for all of them to be using the reverse strand.

ArmandoR1b
01-22-2017, 01:49 PM
Thanks ArmandoR1b, the SNP should have read CTS1751, (my typo put CTS11751) CTS1751 is under R-L21, DF13, Z38589

Here are his results;

SNP Ancestral (Negative) Derived (Positive)
L80 A G -- not found
L157 T A -- TT
M253 C T -- CC
L187 A T -- AA
P30 G A -- not found
L845 T G -- TT

L840 GG -- GG

SNP Ancestral (Negative) Derived (Positive)
P310 A C -- CC
L52 C T -- TT
L51 G A -- AA
L21 C G -- GG
CTS5396 T C -- TT
Z249 T C -- TT
S5668 A G -- AA
FGC13499 G A -- CC
Z16332 A G -- AA
A6 C T -- not found

His CTS1751 is CC. I thought it should be A-G
S7123 is CC

I will wait for his Big y results. This is the 2nd time FTDNA have changed his haplogroup.
So you know that he is definitely L21. DF13 and it's equivalents aren't tested by Geno 2.0 NG and neither is Z39589 so there is no way to determine if he really belongs in the CTS1751 or DF13 subclades. He definitely is not S7123+ because of the following:

CTS5396 T C -- TT
Z249 T C -- TT
S5668 A G -- AA
FGC13499 G A -- CC


CTS1751 is A-G and all of the Geno 2.0 NG kits that I have are A which means that they are negative for CTS1751. So in all of those kits they have a good result. However, S7123 is C for all of the kits that have S7123 in their results. That SNP should not have been used and the upstream SNPs should have been used because they are negative.