PDA

View Full Version : Am I really J-M67?



Flub
03-30-2022, 07:37 PM
So, I've been assigned J-M67 on 23andme, I saw it wasn't specific enough for me so, I went ahead and put some research on it.
I've figured MorleyDna and Yseq subclade predictors could assist on untested subclades that 23andme had missed. What suprised me wasn't that I tested positive for M67/S51, but I was tested negative for a vital connector SNP for it (CTS2130/PF5087) on both MorleyDna and Yseq.

The SNP phylogeny: M172+ ---> M410+ ---> L26+ ---> PF5087- ---> M67+
However, I've also tested positive for L88 which could be the reason why when I tried the Nevgen STR subclade predictor I had J-PF5169 and J-L198 several times for the prediction, but nevertheless 0% probability and 100% unsupported subclade 99.99% of the time.

I don't know if 23andme misses SNPs purposefully especially Ydna SNPs. But, when I've tested full Mtdna sequence on Ftdna they've gave me what I've figured what occured on the jameslink mtdna prediction after I've tested on 23andme. I could see that they've both came close to the same. Jameslink and Ftdna both highlighted the negative mutations for my overall haplogroup, which can be an entirely different subclade from the one they've predicted.
On Ftdna I've joined the "J2-M172 group". None really matched my STR values I wouldn't be suprised because I only had 2 matches come up. But, now I've upgraded to the Big Y and just waiting to see what it shows afterwards.

Gentica277282
03-30-2022, 07:51 PM
Yes you are part of greatness

Flub
03-30-2022, 10:11 PM
Yes you are part of greatness

Yeah, it seems cool. But, I'm skeptical about it. It would seem like it happens to be another cluster in J2a like me and others who are on my relative list are apart of. Perhaps one that's restricted to the area. We'll just have to wait and see.

Flub
04-02-2022, 05:15 PM
I'd like to mention that I've also tested negative for all the SNPs downstream. So, really my group could be indeed rare or there was a mistake done(I doubt it). ---> M172+ ---> M410+ ---> L26+ ---> PF5087- ---> M67+ ---> (?). Again, 2 matches. None of them being on the 67 or 111 on Ftdna, but Y37 to Y12 and all from different regions.

ArmandoR1b
04-02-2022, 08:08 PM
So, I've been assigned J-M67 on 23andme, I saw it wasn't specific enough for me so, I went ahead and put some research on it.
I've figured MorleyDna and Yseq subclade predictors could assist on untested subclades that 23andme had missed. What suprised me wasn't that I tested positive for M67/S51, but I was tested negative for a vital connector SNP for it (CTS2130/PF5087) on both MorleyDna and Yseq.

The SNP phylogeny: M172+ ---> M410+ ---> L26+ ---> PF5087- ---> M67+
However, I've also tested positive for L88 which could be the reason why when I tried the Nevgen STR subclade predictor I had J-PF5169 and J-L198 several times for the prediction, but nevertheless 0% probability and 100% unsupported subclade 99.99% of the time.

I don't know if 23andme misses SNPs purposefully especially Ydna SNPs. But, when I've tested full Mtdna sequence on Ftdna they've gave me what I've figured what occured on the jameslink mtdna prediction after I've tested on 23andme. I could see that they've both came close to the same. Jameslink and Ftdna both highlighted the negative mutations for my overall haplogroup, which can be an entirely different subclade from the one they've predicted.
On Ftdna I've joined the "J2-M172 group". None really matched my STR values I wouldn't be suprised because I only had 2 matches come up. But, now I've upgraded to the Big Y and just waiting to see what it shows afterwards.

23andme never vetted false negatives and false positives for the v5 test. 23andme ignores negatives and just uses positives so it ignores false negatives if there is a downstream positive even if the positive is a false positive.

CTS2130/PF5087 is possibly a false negative for everyone that is positive for M67. I can assure you that M67 is not a false positive because I have looked at dozens of 23andme kits and no one that isn't part of the J2 haplogroup has that as a positive result. I have identified 9 false positives in 23andme v5 but none of them are in the J2 haplogroup.

False negatives are hard to identify but for M67 we just need about 5-10 other M67, or downstream, people to post their 23andme v5 results which would make it evident how common the PF5087- appears for people that are positive for M67 so we can see what they got for PF5087 which is hg19 position 7284302 and is rs371285248 which can be looked up at https://you.23andme.com/tools/data/?query=rs371285248 I am sure that will always be negative since I have never heard of a person tested through 23andme v5 with a result of J-PF5087 for their paternal haplogroup.

L88.1 is a C to T mutation. L88.2 is a T to C mutation. Both are at hg19 position 17595842. It is a position that changes frequently and is unreliable. You can check your result at https://you.23andme.com/tools/data/?query=rs3966071 we should check this with other R-M67 23andme v5 people also.

Since L88 hg19 position 17595842 is unreliable and PF5087 is almost surely a false negative and M67 is not a false positive that only leads to one conclusion. You belong to haplogroup M67.

Y-DNA and mtDNA are very different because of the massively higher number of mutations in Y-DNA than in mtDNA so they shouldn't be compared like you did.

What is the same between Y-DNA and mtDNA at 23andme is that they both test a limited number of positions and there are more accurate tests elsewhere such as Big Y for Y-DNA and Full mtDNA sequencing for mtDNA or WSG at Yseq.

It's a big myth that MorleyDna and Yseq subclade predictors could assist on untested subclades that 23andme had missed. MorleyDna and Yseq don't do additional testing and therefore can't find more downstream SNPs that weren't reported by 23andme. I have no idea why that myth persists and why I see so many people thinking that they help identify subclades not reported by 23andme. I'd like to see just one person with a 23andme v5 test that found a more specific haplogroup that Yseq cladefinder has found that 23andme didn't. The only way that you can find additional SNPs downstream from what 23andme reports is by getting advanced testing.

I still advocate using Yseq Cladefinder subclade predictor so that people can see what wasn't tested by 23andme. However, a lot of people have a hard time grasping what raw DNA is and how 23andme tests work and how NGS and WGS tests work or that the blue SNPs in the Yseq Cladefinder subclade predictor are SNPs not tested by 23andme.

You did the right thing in getting the Big Y test since very few tests beat it.

Flub
04-02-2022, 08:34 PM
23andme never vetted false negatives and false positives for the v5 test. 23andme ignores negatives and just uses positives so it ignores false negatives if there is a downstream positive even if the positive is a false positive.

CTS2130/PF5087 is possibly a false negative for everyone that is positive for M67. I can assure you that M67 is not a false positive because I have looked at dozens of 23andme kits and no one that isn't part of the J2 haplogroup has that as a positive result. I have identified 9 false positives in 23andme v5 but none of them are in the J2 haplogroup.

False negatives are hard to identify but for M67 we just need about 5-10 other M67, or downstream, people to post their 23andme v5 results which would make it evident how common the PF5087- appears for people that are positive for M67 so we can see what they got for PF5087 which is hg19 position 7284302 and is rs371285248 which can be looked up at https://you.23andme.com/tools/data/?query=rs371285248 I am sure that will always be negative since I have never heard of a person tested through 23andme v5 with a result of J-PF5087 for their paternal haplogroup.

L88.1 is a C to T mutation. L88.2 is a T to C mutation. Both are at hg19 position 17595842. It is a position that changes frequently and is unreliable. You can check your result at https://you.23andme.com/tools/data/?query=rs3966071 we should check this with other R-M67 23andme v5 people also.

Since L88 hg19 position 17595842 is unreliable and PF5087 is almost surely a false negative and M67 is not a false positive that only leads to one conclusion. You belong to haplogroup M67.

Y-DNA and mtDNA are very different because of the massively higher number of mutations in Y-DNA than in mtDNA so they shouldn't be compared like you did.

What is the same between Y-DNA and mtDNA at 23andme is that they both test a limited number of positions and there are more accurate tests elsewhere such as Big Y for Y-DNA and Full mtDNA sequencing for mtDNA or WSG at Yseq.

It's a big myth that MorleyDna and Yseq subclade predictors could assist on untested subclades that 23andme had missed. MorleyDna and Yseq don't do additional testing and therefore can't find more downstream SNPs that weren't reported by 23andme. I have no idea why that myth persists and why I see so many people thinking that they help identify subclades not reported by 23andme. I'd like to see just one person with a 23andme v5 test that found a more specific haplogroup that Yseq cladefinder has found that 23andme didn't. The only way that you can find additional SNPs downstream from what 23andme reports is by getting advanced testing.

I still advocate using Yseq Cladefinder subclade predictor so that people can see what wasn't tested by 23andme. However, a lot of people have a hard time grasping what raw DNA is and how 23andme tests work and how NGS and WGS tests work or that the blue SNPs in the Yseq Cladefinder subclade predictor are SNPs not tested by 23andme.

You did the right thing in getting the Big Y test since very few tests beat it.

Thanks for the explanation.

Flub
04-06-2022, 06:22 AM
23andme never vetted false negatives and false positives for the v5 test. 23andme ignores negatives and just uses positives so it ignores false negatives if there is a downstream positive even if the positive is a false positive.

CTS2130/PF5087 is possibly a false negative for everyone that is positive for M67. I can assure you that M67 is not a false positive because I have looked at dozens of 23andme kits and no one that isn't part of the J2 haplogroup has that as a positive result. I have identified 9 false positives in 23andme v5 but none of them are in the J2 haplogroup.

False negatives are hard to identify but for M67 we just need about 5-10 other M67, or downstream, people to post their 23andme v5 results which would make it evident how common the PF5087- appears for people that are positive for M67 so we can see what they got for PF5087 which is hg19 position 7284302 and is rs371285248 which can be looked up at https://you.23andme.com/tools/data/?query=rs371285248 I am sure that will always be negative since I have never heard of a person tested through 23andme v5 with a result of J-PF5087 for their paternal haplogroup.

L88.1 is a C to T mutation. L88.2 is a T to C mutation. Both are at hg19 position 17595842. It is a position that changes frequently and is unreliable. You can check your result at https://you.23andme.com/tools/data/?query=rs3966071 we should check this with other R-M67 23andme v5 people also.

Since L88 hg19 position 17595842 is unreliable and PF5087 is almost surely a false negative and M67 is not a false positive that only leads to one conclusion. You belong to haplogroup M67.

Y-DNA and mtDNA are very different because of the massively higher number of mutations in Y-DNA than in mtDNA so they shouldn't be compared like you did.

What is the same between Y-DNA and mtDNA at 23andme is that they both test a limited number of positions and there are more accurate tests elsewhere such as Big Y for Y-DNA and Full mtDNA sequencing for mtDNA or WSG at Yseq.

It's a big myth that MorleyDna and Yseq subclade predictors could assist on untested subclades that 23andme had missed. MorleyDna and Yseq don't do additional testing and therefore can't find more downstream SNPs that weren't reported by 23andme. I have no idea why that myth persists and why I see so many people thinking that they help identify subclades not reported by 23andme. I'd like to see just one person with a 23andme v5 test that found a more specific haplogroup that Yseq cladefinder has found that 23andme didn't. The only way that you can find additional SNPs downstream from what 23andme reports is by getting advanced testing.

I still advocate using Yseq Cladefinder subclade predictor so that people can see what wasn't tested by 23andme. However, a lot of people have a hard time grasping what raw DNA is and how 23andme tests work and how NGS and WGS tests work or that the blue SNPs in the Yseq Cladefinder subclade predictor are SNPs not tested by 23andme.

You did the right thing in getting the Big Y test since very few tests beat it.


What I don't get however though is; why miss a SNP? 23&me should be testing these SNPs.

ArmandoR1b
04-07-2022, 01:48 AM
What I don't get however though is; why miss a SNP? 23&me should be testing these SNPs.

I'm not sure what you are asking. What do you consider a missed SNP?

23andme v5 uses an Illumina GSA chip that tests specific SNPs. The chips can test multiple people at a time. This keeps cost down. If they tested more SNPs then testing would be slower and more costly. If they charged more then most people would go with a competitor. So some SNPs aren't tested due to a cost-benefit analysis. So those aren't misses. They are just not tested.

As you saw in another thread almost everyone had the same SNP reported by 23andme that was reported by the Yseq Cladefinder tool. There was only one and we still aren't sure what happened with that one because other testing hasn't been done. So there really aren't SNPs that are tested by 23andme that aren't reported by 23andme unless they are false negatives but those should just not be reported as tested. That is what vetting is supposed to do. So technically those aren't really misses either.

Flub
04-07-2022, 02:59 AM
I'm not sure what you are asking. What do you consider a missed SNP?

23andme v5 uses an Illumina GSA chip that tests specific SNPs. The chips can test multiple people at a time. This keeps cost down. If they tested more SNPs then testing would be slower and more costly. If they charged more then most people would go with a competitor. So some SNPs aren't tested due to a cost-benefit analysis. So those aren't misses. They are just not tested.

As you saw in another thread almost everyone had the same SNP reported by 23andme that was reported by the Yseq Cladefinder tool. There was only one and we still aren't sure what happened with that one because other testing hasn't been done. So there really aren't SNPs that are tested by 23andme that aren't reported by 23andme unless they are false negatives but those should just not be reported as tested. That is what vetting is supposed to do. So technically those aren't really misses either.

Pardons, I'll rephrase it. 'Why mark the SNP negative when supposedly it really is positive...?'

ArmandoR1b
04-08-2022, 02:45 AM
Pardons, I'll rephrase it. 'Why mark the SNP negative when supposedly it really is positive...?'

Because they didn't vet it. The test did find it negative but the test is designed in a way that it doesn't correctly read some SNPs. A vetting process should find customers that are positive for downstream SNPs, that aren't positive for every single tester regardless of haplogroup, yet have a negative SNP that should be positive. Once those are found they should have determined that it is a false negative and remove the reporting of the SNP completely since the test is flawed for that SNP.

The failure to include a vetting process caused the false reporting of the SNP as negative.