A problem with the "Eddowes Shawl" DNA match

[tji]

Hi Deb's

I understood the alignment position to be 310 T and then 5 cytosines to position 315 in the CRS? Is that wrong?

[/QUOTE]

If I am understanding correctly the .1C after the 315 denotes the mutation designation point - so the difference to the CRS database is 1 added cytosine, the .2 would be 2 added cytosines and so on. It also makes it clear that it is an added mutation not a substitute from one of the other 3 bases (adenine, thymine, and guanine)

I am not sure if any of that makes sense to anyone else but I think I understand it

Tracy

[Debra A]

Hi Deb's



If I am understanding correctly the .1C after the 315 denotes the mutation designation point - so the difference to the CRS database is 1 added cytosine, the .2 would be 2 added cytosines and so on. It also makes it clear that it is an added mutation not a instead of substitute from one of the other 3 bases (adenine, thymine, and guanine)

I am not sure if any of that makes sense to anyone else but I think I understand it

Tracy

Hi Tracy
Thanks, yes, that's how I understand it too.
I was wondering where Pete was starting the C count from.

[Theagenes]

I just want to thank everyone who brought this out and and are trying to determine what it means for Jari's resluts. After reading the original parer and several others on the problems of nomenclature with polycytosine regions like this, i think I'm starting to get a handle on it and I'm following what Debra is trying to say.

This one is relevent I believe. Again, it's discussing how another software program deals with the nomenclature error and the rules of nomenclature for problem regions like this.

http://mitotech.net/files/Mitotyper_Description_Paper_9-15-08%20final.pdf

On page 8 it begins discussing the our Poly-C region and explains the issue with the rCRS only showing 5 Cs instead of 6 that Debra was talking about.

But from doing a bit of searching around it does seem like this is a "known" issue. It's hard to believe that Jari, wouldn't have been aware that a 314.1c is just a misreporting of 315.1c. Maybe he has old software and didn't know this was a problem? I can't believe that once he got this supposed rare result he didn't go to the literature immediately, where he would seen this was an issue.

Or is that the shawl and Miller DNA are part of that rare group that only has 5 Cs in the region instead of 6 that appears in the rCRS? And Edwards garbled this somehow? I really want to hear what Jari has to say about this before rushing to any judgment.

[Debra A]

Thanks for the link to that paper Theagenes...that's what I meant!
I had exactly the same reservations about it being an error initially too , given Jari's experience and the fact that it is a known issue and software is specially designed to overcome mistakes like this, but I do think Chris is right to highlight this and I hope JL will address the questions about it at some point. He has been asked, I believe.

[Theagenes]

Thanks for the link to that paper Theagenes...that's what I meant!
I had exactly the same reservations about it being an error initially too , given Jari's experience and the fact that it is a known issue and software is specially designed to overcome mistakes like this, but I do think Chris is right to highlight this and I hope JL will address the questions about it at some point. He has been asked, I believe.

I've found a number of similar articles addressing this, as well as software patents describing how their product deals with it. It appears that this is the article that first identified the issue (i.e. that the 5 C sequence in the CRS was the exception, not the rule):

Andrews et al., “Reanalysis And Revision Of The Cambridge Reference Sequence For Human Mitochondrial DNA” Nature Genetics 23, 147 (1999)

http://www.nature.com/ng/journal/v23/n2/full/ng1099_147.html

I should have access to the full article at work, so I'll try to get a copy, but it does seem that this has been known for some time.

One thing I can't find is what the proper nomenclature is for describing that rare sequence that really does only have five Cs in that region instead of the normal 6. If that is what the shawl/Miller results show that might explain Jari describing this as such as rare mutation, while still being quite aware of the typical problems with this particular poly-C region. In other words, maybe the shawl/Miller DNA has the rare 5 cytosine instead of 6 sequence that caused this problem in the CRS to begin with.

Is is possible that 314.1c is one way of describing this? That is, 311 through 314 are all Cs, plus one extra C (total 5), with 315 empty (where the 6th C should normally be). If there is no accepted way of denoting this because it is so rare then how would you describe it if you acually got it as a result? Especially since most software it seems automatically changes a 314.1c result to 315.1c.

And if Jari tried to describe all this to Edwards -- good lord!

And do think that just as most people were careful to use lots of qualifers when this initially sounded promising, we should also be carefull not to throw everything out just yet either. But we really need to hear Jari's response to this.

[mickreed]

And if Jari tried to describe all this to Edwards -- good lord!

You speak truly, Theagenes.

Isn't there an 'real expert' web site where we might scrounge some help with this, rather than us trying to find and interpret papers? From what we've been hearing Jari has been advised to keep his head in, so it could be a long time before we hear anything definitive from him - if ever.

[Robert]

Lars says Jari's going to be on TV :

http://www.jtrforums.com/showthread.php?t=22254&page=69

Maybe it's Come Dnacing (sic)

[Peter Griffith aka gryff]

In other words, maybe the shawl/Miller DNA has the rare 5 cytosine instead of 6 sequence that caused this problem in the CRS to begin with.

....

But we really need to hear Jari's response to this.

You make a very good point Theagenes.

You speak truly, Theagenes.

From what we've been hearing Jari has been advised to keep his head in, so it could be a long time before we hear anything definitive from him - if ever.

By whom mick? The university, Edwards, the publisher?

And if there is no response ever - does that not look bad too?

cheers, gryff

[Fisherman]

And if there is no response ever - does that not look bad too?

cheers, gryff

Not for the sales figures on Edwards book, it don´t.

Otherwise, though ...

The best,
Fisherman

[Theagenes]

You speak truly, Theagenes.

Isn't there an 'real expert' web site where we might scrounge some help with this, rather than us trying to find and interpret papers? From what we've been hearing Jari has been advised to keep his head in, so it could be a long time before we hear anything definitive from him - if ever.

The problem is that even if we got someone knowledgeable in this area to help, we have no real information to give them. Everything we have except a few snippets from Jari's interview comes filtered through Edwards' book, and now that you've read it as well, you know how big a problem that is.

At first it seemed as though this supposed rare mutation was something Edwards was unlikely to have misunderstood -- either it's there or it isn't. But now it could be that Jari misinterpreted what is clearly a problematic area. Or he understood the problems perfectly well and tried to explain it to Edwards but the latter didn't understand or heard only what he wanted to hear -- who knows at this point.

But there was a recent poster in the other thread whose wife was a geneticist and has been following this. Maybe she can chime in.

[Debra A]

One thing I can't find is what the proper nomenclature is for describing that rare sequence that really does only have five Cs in that region instead of the normal 6. If that is what the shawl/Miller results show that might explain Jari describing this as such as rare mutation, while still being quite aware of the typical problems with this particular poly-C region. In other words, maybe the shawl/Miller DNA has the rare 5 cytosine instead of 6 sequence that caused this problem in the CRS to begin with.

Is is possible that 314.1c is one way of describing this? That is, 311 through 314 are all Cs, plus one extra C (total 5), with 315 empty (where the 6th C should normally be). If there is no accepted way of denoting this because it is so rare then how would you describe it if you acually got it as a result? Especially since most software it seems automatically changes a 314.1c result to 315.1c.

And if Jari tried to describe all this to Edwards -- good lord!

And do think that just as most people were careful to use lots of qualifers when this initially sounded promising, we should also be carefull not to throw everything out just yet either. But we really need to hear Jari's response to this.

Would it be reported at all if the Miller/shawl DNA was the rare 5 c sequence I wonder? Are only differences from the reference sequences noted? I read the CRS is technically a deletion at this region but as it is the reference sequence the nomenclature has not been changed to reflect this.

[Theagenes]

Would it be reported at all if the Miller/shawl DNA was the rare 5 c sequence I wonder? Are only differences from the reference sequences noted? I read the CRS is technically a deletion at this region but as it is the reference sequence the nomenclature has not been changed to reflect this.

Good point!

[Chris]

Chris is being entirely too humble here. I was lucky enough to find the information in question, but it is he who spent the time deciphering and understanding it all.

I am posting this because it is not something that is easy to come by and he has spent a lot of time on this and so I think regardless of whether people think he is right or wrong he should be credited for his perseverance .

Thanks Tracy

But perhaps you are being a bit too humble yourself. After all, the key points about 314.1C being an error in nomenclature for 315.1C, and 315.1C being common, are explicitly there in what you posted.

Mick Reed has pointed out that he had also posted a link to the same paper, independently of Tracy.

I'd also like to thank Tracy, Debs, Pat and Alex for private discussions over the last week or so.

[Chris]

No, the paper is nothing to do with identifying genes at all - according to your post it's a - "paper describing software designed to identify missing persons" and what the paper explains is that the software cannot distinguish between these two particular sequences

The paper is not claiming that these mutations are identical, just that the software cannot tell the difference between 314.1c and 315.1c

Yes. The software is designed to identify missing persons. But it does that partly by matching DNA. We should hardly be discussing it otherwise.

It does not say that the software "cannot distinguish between these two particular sequences". It says the software "tolerates errors in nomenclature for equivalent variants". That means it allows for the fact that people may be calling the same sequence by two different names.

I haven't 'miscounted' anything I had copied your example.

The 'Poly-cytosine region' is the chain of C's - 314.1C is located there
'After the base position' is after the G - 315.1C is located there

They are all "base positions". The nucleotide in position 315 in the reference sequence is the final C in the string of five. Therefore inserting a C after position 315 gives TCCCCCCG, not TCCCCCGC. TCCCCCGC would be the result of inserting a C after position 316.

[Chris]

One thing I can't find is what the proper nomenclature is for describing that rare sequence that really does only have five Cs in that region instead of the normal 6. If that is what the shawl/Miller results show that might explain Jari describing this as such as rare mutation, while still being quite aware of the typical problems with this particular poly-C region. In other words, maybe the shawl/Miller DNA has the rare 5 cytosine instead of 6 sequence that caused this problem in the CRS to begin with.

Having five Cs there is just referred to as 311C-315C in this discussion of rare polymorphisms in the reference sequence:
"Seven nucleotides are considered to be rare polymorphisms and were determined to be correct as originally sequenced (J01415 gi:337188). Nucleotides 263A, 311C-315C, 750A, 1438A, 4769A, 8860A, and 15326A are considered to be rare polymorphisms and are maintained as part of the true reference sequence."

http://www.mitomap.org/MITOMAP/HumanMitoSeq

I don't think it would make sense to describe 5 Cs as the insertion of a C, at whatever position. The other thing is that having five Cs is rare, but nowhere near as rare as 1 in 290,000. The figure I quoted above would suggest something like 1%, but it may be commoner than that in the UK because the reference sequence was based on someone from Britain.

As Debs suggests, it would make sense to report agreement with the reference sequence, where the reference sequence contains a rare polymorphism. At least the EMPOP database (and probably others) takes care of that automatically - if you put a range of 315.1 and nothing in the right-hand box, it will give only a small number of exact matches, reflecting the rare condition of only five Cs.