A problem with the "Eddowes Shawl" DNA match

[Mr Lucky]

The more of this stuff you come out with, the clearer you make it that you don't understand the first thing about the topic of discussion.

What I'm trying to understand is why you think putting the wrong information into the wrong data base and noting the wrong answer comes out means that there is a problem with a match between two DNA samples

Why would that ever be the case?

[Chris]

What I'm trying to understand is why you think putting the wrong information into the wrong data base and noting the wrong answer comes out means that there is a problem with a match between two DNA samples

This is the most astonishing behaviour I've ever witnessed on Casebook. And after 15 years, that's saying something!

[Mr Lucky]

Oh dear! I will try to explain

1) those databases just give you a statistical likelihood of the mutation occurring.

2) They don't tell you whether two different DNA sample match or not

3) nothing you can type into them will ever demonstrate that the two samples don't match.

That's the simplest way that it can be explained.

So going back to main point have you got any evidence to back up your claim that there is a "problem with the Eddowes shawl DNA match"

[pinkmoon]

Oh dear! I will try to explain

1) those databases just give you a statistical likelihood of the mutation occurring.

2) They don't tell you whether two different DNA sample match or not

3) nothing you can type into them will ever demonstrate that the two samples don't match.

That's the simplest way that it can be explained.

So going back to main point have you got any evidence to back up your claim that there is a "problem with the Eddowes shawl DNA match"

What about the fact that there is no evidence to link that shawl to any of the murders and the shawl itself can't even be dated to 1888.

[Chris]

Mr Lucky

I've really don't think I've ever encountered anyone with as little grasp of the subject they are trying to argue about.

The point, obviously, is that Dr Louhelainen was quoted as having obtained from this database a frequency of 1 in 290,000 for the "314.1C" mutation. That's why I performed a search in the database for the interpretation you suggested - clearly mistaken though you were a priori - in order to demonstrate that its frequency in the database was nothing like that figure.

But you are just trolling now, aren't you?

[mickreed]

Oh dear! I will try to explain

1) those databases just give you a statistical likelihood of the mutation occurring.

2) They don't tell you whether two different DNA sample match or not

3) nothing you can type into them will ever demonstrate that the two samples don't match.

That's the simplest way that it can be explained.

So going back to main point have you got any evidence to back up your claim that there is a "problem with the Eddowes shawl DNA match"

Mr Lucky

This seems rather futile. Go back to post 1 in this thread and see why Chris started it. There really does seem to be a problem with the way in which JL described the issue to RE and which is put out in the book. The article that Tracy and I independently came across contained sufficient info to cause us to wonder what JL meant in his description.

That is not the same as saying there is definitely a problem with the match. There does seem to be a problem with the way the match (if that's what it is) is described.

I pretty much said all this a few posts back but you didn't comment (which is fine). You seem only to want to lambast Chris's posts, rather than engage with them sensibly.

I ain't getting involved in this particular bit of nonsense any more, so will just say that until JL clarifies his presentation of the science then very legitimate doubts arise and should be discussed.

[Mr Lucky]

I think your argument that the well known genome variance in the poly-cytosine region is the same thing as the "global private mutation" found in the Eddowes DNA is dead in the water.

I think you better address all the issues you have just side stepped that buries that hypothesis

[mickreed]

I think your argument that the well known genome variance in the poly-cytosine region is the same thing as the "global private mutation" found in the Eddowes DNA is dead in the water.

I think you better address all the issues you have just side stepped that buries that hypothesis

I wasn't going to say any more on this, but Lucky and Chris, can you tell me what this means?

global private mutation: mutation never observed in Phylotree, probably due to inconsistent alignments, phantom mutations or point heteroplasmies (R, Y, K...)

It seems to be saying that a 'global private mutation' is sufficently rare to have never been seen in Phylotree, and is probably due to an analysis 'stuff up'. Is that a fair interpretation? I really don't know but would like to.

[Mr Lucky]

I wasn't going to say any more on this, but Lucky and Chris, can you tell me what this means?

global private mutation: mutation never observed in Phylotree, probably due to inconsistent alignments, phantom mutations or point heteroplasmies (R, Y, K...)

It seems to be saying that a 'global private mutation' is sufficently rare to have never been seen in Phylotree, and is probably due to an analysis 'stuff up'. Is that a fair interpretation? I really don't know but would like to.

Yes, a global private mutation is one not found in databases

Phylotree was the first branching type database IIRC

a lot of supposed global private mutation are caused by errors, inconsistent alignments - exactly what it sounds like. phantom mutations - errors caused by amplification. not sure about 'point heteroplasmies ' but I expect its a type of standard error.

[mickreed]

I wasn't going to say any more on this, but Lucky and Chris, can you tell me what this means?

global private mutation: mutation never observed in Phylotree, probably due to inconsistent alignments, phantom mutations or point heteroplasmies (R, Y, K...)

It seems to be saying that a 'global private mutation' is sufficently rare to have never been seen in Phylotree, and is probably due to an analysis 'stuff up'. Is that a fair interpretation? I really don't know but would like to.

Does this help?

Length heteroplasmy also is observed and typically manifests as variation in the number of bases residing within a homopolymeric stretch (i.e., C stretches). The most commonly occurring form of heteroplasmy (more so that point heteroplasmy) is variation in the number of C’s in a homopolymeric stretch. A number of studies have documented length heteroplasmy within an individual (36-39 ). The mechanism suggested for generating length heteroplasmy is replication slippage (36).

...

A similar region of length heteroplasmy exists in HVII, encompassing positions 303-315. Many individuals exhibit length heteroplasmy in this region. In order to assess the presence of length heteroplasmy, careful consideration should be given to positions 309 and 310 in both light and heavy strand chromatograms. Likewise, should a transition occur in HV II at position 310, no attempt will be made to count the number of cytosine residues in the homopolymeric region for comparison purposes. Length heteroplasmy is exhibited by “out-of- phase sequence carryover”downstream of this region. Therefore, positions after 309 in the light strand and before 309 in the heavy strand should be compared to positions before 309 in the light strand and after 309 in the heavy strand for such evidence.

Again, in most cases, length heteroplasmy is not used for interpretations. However, if one determines it to be of value, a common length variant should be present in both the questioned sample and the known sample in order for a conclusion of sequence concordance to be reached. Also, the amount of length variability in the known sample should be considered in such situations.

Full FBI article here:

https://www.promega.com/~/media/files/resources/conference%20proceedings/ishi%2010/oral%20presentations/37budowle.pdf

[mickreed]

a lot of supposed global private mutation are caused by errors, inconsistent alignments - exactly what it sound like. phantom mutations - errors caused by amplification. not sure about 'point heteroplasmies ' but I expect its a type of standard error.

Thanks Lucky. So, are we not entitled to wonder whether this alleged 'global private mutation' is real or not?

[Mr Lucky]

Thanks Lucky. So, are we not entitled to wonder whether this alleged 'global private mutation' is real or not?

Oh yes, that would be a very interesting line of enquiry.

[Mr Lucky]

another line of enquiry would be trying to understanding the statistical mechanism used to eliminate the errors

[Mr Lucky]

Does this help?

Full FBI article here:
https://www.promega.com/~/media/file.../37budowle.pdf

Yes thanks, very interesting the mtDNA samples from the same individual can have variable C's in the Poly-cytosine region!

The upshot appears to be - avoid errors in sequencing in these areas by matching up both ends of the troublesome area at the same time.

This is some thing you may want to look into further from your perspective

[RockySullivan]

Are you admitting there isn't a problem with the DNA match?

What match?