A6 Murder DNA evidence

[Victor]

Hi Vic
Points 1 to 4

1) You're welcome. Rather you than me mate

Didn't happen, another week living out of boxes and suitcases - hopefully moving next weekend now!

2) I'm not sure what you mean here, it seems like a contradiction. Even so I stick by what I put.

Contamination is not definite. If it was definite then you could predict it and account for it and therefore it wouldn't be an issue.

It is an issue because it is unpredictable. There is a probability associated with contamination and the control is there because the probability of the control being contaminated at the same time as the sample is very large.

In other words the existance of the blank control is further evidence that contamination is probability based.

3) The 70% figure refers to the number of DNA analysis tests that should be abandoned due to spoiling in the blank sample and I got that from The Forensic Institute in Glasgow.

OK, so 70% of the time the blank control is contaminated. One could infer that the probability of contamination is therefore 70%

The 6% figure refers to the number of DNA analysis results in one calender year that could actually be used by one of her Maj's Old Bill county forces to even contemplate using as evidence. This was gleaned from a report I read after the Hoey appeal.

94% of DNA analyses are not used in evidence, but that says nothing about whether those analyses were useful or not. Whether they were mixed profiles and therefore inconclusive. Whether they only gave results of the victim's DNA. Whether they gave negative results.

Where's the comparable data for another technique? For example:-
What proportion of fingerprint analyses are used in evidence?
How many fingerprint analyses show no fingerprints?

No I didn't know that. But what does that tell you?

That tells me that statistics can be massaged to suit any situation, and that the same statistics can be used to prove opposite sides of the same argument...for example at University we were asked to look at the case of Gregor Mendel and his work on genetic inheritance in pea plants - some have argued that his results are too accurate and are therefore fraudulent.

4) I take your point. But the DNA techniques we are talking about will not solve anything unless other good evidence is also present. If the fuzz can get a confession, legitimately, from a suspect using DNA analysis to say they were at a crime scene then fair enough.

Regards
Reg

It's a bit difficult to get a confession from a dead man! But that's just argumentative, what I think you are saying is that there needs to be corroborative evidence, like VS's identification, and that all the evidence should be considered as a whole, whereas the Hanratty appeal only considered NEW evidence like the DNA and non-disclosure of materials to the original defence team.

KR,
Vic.

Cheers Reg.



Categorically contamination is dependant upon probability, otherwise you could predict the contamination and account for it and disregard it.



What's the source for these figures? I've seen similar figures but for different things, does the 70% contaminated analyses include the ones where they are just chancing their arm in the hope of possibly getting a profile, but realistically don't expect to get one?

Lies, damn lies and statistics! That 6% figure could actually be successful convictions in cases where DNA has been used - even if it is used to prove someone wasn't guilty!

Did you know that police statistics include as a successful conviction those cases where the court finds someone not guilty? The criminal isn't identified, but it's a resolved crime.



I can offer a different answer - So that murdering scumbags are taken off the streets and aren't free to rape and kill other people. In other words - to save lives and minimise anguish.

Of course peoples viewpoint can change when they actually have a partner or family member that has been raped or assaulted or murdered - then they tend to want the police to use every possible method to try and work out who the perpetrator was.

KR,
Vic.

Hi Vic
Points 1 to 4

1) You're welcome. Rather you than me mate

2) I'm not sure what you mean here, it seems like a contradiction. Even so I stick by what I put.

3) The 70% figure refers to the number of DNA analysis tests that should be abandoned due to spoiling in the blank sample and I got that from The Forensic Institute in Glasgow.

The 6% figure refers to the number of DNA analysis results in one calender year that could actually be used by one of her Maj's Old Bill county forces to even contemplate using as evidence. This was gleaned from a report I read after the Hoey appeal.

No I didn't know that. But what does that tell you?

4) I take your point. But the DNA techniques we are talking about will not solve anything unless other good evidence is also present. If the fuzz can get a confession, legitimately, from a suspect using DNA analysis to say they were at a crime scene then fair enough.

Regards
Reg

[Victor]

Hi Vic
Good luck with the move.

Cheers Reg.

Contamination is not a probability game within DNA testing. Contamination is assumed to be there. Thats why control samples are run in parallel.

Categorically contamination is dependant upon probability, otherwise you could predict the contamination and account for it and disregard it.

Advanced PCR testing strategies such as LCN/LT for instance return about 70% contaminated analyses and must be discarded. Some police forces get successful return rates as low as 6% meaning that the failure rate for that forces particular destructive DNA forensic testing was 94%. Confusing? uummmmm!

What's the source for these figures? I've seen similar figures but for different things, does the 70% contaminated analyses include the ones where they are just chancing their arm in the hope of possibly getting a profile, but realistically don't expect to get one?

Lies, damn lies and statistics! That 6% figure could actually be successful convictions in cases where DNA has been used - even if it is used to prove someone wasn't guilty!

Did you know that police statistics include as a successful conviction those cases where the court finds someone not guilty? The criminal isn't identified, but it's a resolved crime.

This doesn't take into account the delivered results that have actually been analysed incorrectly (knowingly or otherwise).
If these advanced PCR techniques are so poor in returning decent results what is the reason they are being used as evidence in the criminal justice system.
I can answer that right now. Your average lay man (including 3 well known court of appeal judges) does not know the first thing about DNA technologies. The CSI effect kicks in and juries suddenly expect and rely on DNA evidence to be able to convict without any sense of getting the verdict wrong.

Reg

I can offer a different answer - So that murdering scumbags are taken off the streets and aren't free to rape and kill other people. In other words - to save lives and minimise anguish.

Of course peoples viewpoint can change when they actually have a partner or family member that has been raped or assaulted or murdered - then they tend to want the police to use every possible method to try and work out who the perpetrator was.

KR,
Vic.

Hi DM,

That's the thing about contamination, it's a probability game. You can have contamination even with proper techniques, and conversely you can have no contamination even with sloppy techniques.

Good techniques reduce the probability but doesn't eliminate contamination
Poor techniques increase the probability but doesn't guarantee contamination

Statistics and probability are confusing, even if you do understand Schrodinger's Cat.

KR,
Vic.

Hi Vic
Good luck with the move.

Contamination is not a probability game within DNA testing. Contamination is assumed to be there. Thats why control samples are run in parallel.
Advanced PCR testing strategies such as LCN/LT for instance return about 70% contaminated analyses and must be discarded. Some police forces get successful return rates as low as 6% meaning that the failure rate for that forces particular destructive DNA forensic testing was 94%. Confusing? uummmmm!
This doesn't take into account the delivered results that have actually been analysed incorrectly (knowingly or otherwise).
If these advanced PCR techniques are so poor in returning decent results what is the reason they are being used as evidence in the criminal justice system.
I can answer that right now. Your average lay man (including 3 well known court of appeal judges) does not know the first thing about DNA technologies. The CSI effect kicks in and juries suddenly expect and rely on DNA evidence to be able to convict without any sense of getting the verdict wrong.

Reg

[Victor]

Hi DM,

That's the thing about contamination, it's a probability game. You can have contamination even with proper techniques, and conversely you can have no contamination even with sloppy techniques.

Good techniques reduce the probability but doesn't eliminate contamination
Poor techniques increase the probability but doesn't guarantee contamination

Statistics and probability are confusing, even if you do understand Schrodinger's Cat.

KR,
Vic.

[Dupplin Muir]

Hello Victor

It's funny how we can read different things into the same piece of text.

The number of qualifiers used is astounding...(last sentence of the first paragraph DM quoted post #111)
"Contamination can influence PCR results, particularly in the absence of proper handling techniques and proper controls for contamination."

I took that 'particularly' to mean that you can get contamination EVEN WITH 'proper handling techniques and proper controls for contamination'!

I can only conclude that it's a glass-half-full/glass-half-empty kind of thing!

DM

[Victor]

Hi Vic
What did you make of the rest of the 2 articles that DM posted? Are you just cherry picking the bits that make you feel good? Ostrichs bury their heads in the sand too!

Reg

Pot, Kettle, Colour-check!

No, I haven't got round to reading the other article yet.

I did think that discovering a technique that would explain the "contamination would have to be semen" comment that has been a point of contention for quite a while was significant, whereas lots of "extreme care needed to minimise contamination" type comments are futile if there were no contaminant profiles detected.

The number of qualifiers used is astounding...(last sentence of the first paragraph DM quoted post #111)
"Contamination can influence PCR results, particularly in the absence of proper handling techniques and proper controls for contamination."

KR,
Vic.

Hi DM,

I've just read through this link and it does contain some very relevant points - in particular the part where it describes centrifuging the sample to seperate the sperm and non-sperm DNA so that a DNA profile of just the sperm is possible.

Now that really is very interesting if it was used in the A6 case!

KR,
Vic.

Hi Vic
What did you make of the rest of the 2 articles that DM posted? Are you just cherry picking the bits that make you feel good? Ostrichs bury their heads in the sand too!

Reg

[Victor]

Hi Vic

Whether it was used or not it still does explain how no other profiles were found on the hanky, considering how many persons may have handled it!

Reg

Hi Reg,

Yes but it does very nicely explain why any contamination on the knicker fragment would have to be semen! Because the DNA profiles came from the semen-only fraction.

And of course your qualifier I've highlighted above explains the hanky perfectly well.

KR,
Vic.

Hi DM,

I've just read through this link and it does contain some very relevant points - in particular the part where it describes centrifuging the sample to seperate the sperm and non-sperm DNA so that a DNA profile of just the sperm is possible.

Now that really is very interesting if it was used in the A6 case!

KR,
Vic.

Hi Vic

Whether it was used or not it still does explain how no other profiles were found on the hanky, considering how many persons may have handled it!

Reg

[Victor]

Here are another couple of interesting sites and I've extracted some points that seem relevant to the A6 case.

http://www.scientific.org/tutorials/...ley/riley.html

Hi DM,

I've just read through this link and it does contain some very relevant points - in particular the part where it describes centrifuging the sample to seperate the sperm and non-sperm DNA so that a DNA profile of just the sperm is possible.

Now that really is very interesting if it was used in the A6 case!

KR,
Vic.

[Victor]

Hello all

I finally got around to watching the 2002 'Horizon' on the A6 murder, and I fear that it rather fell into its usual trap of 'trust the men in the white coats'. I was also rather surprised that it used the false logic of asking 'if JH wasn't the rapist, where's the rapist's DNA?'

I believe this was mentioned on the old thread, but one counter to this argument would be that the rapist was azoospermic. The website

http://www.andrologyjournal.org/cgi/content/full/23/3/293

puts the incidence of this condition at 2%. Given some of the improbabilities and coincidences surrounding this case, odds of 50:1 seem like a racing cert!

DM (not FSS: that was me playing around with the post in Notepad and not proofreading properly!)

Hi DM,

I'll reply to your other post later, but to comment on the above point, it was established that MG is blood group AB, and they detected some blood group O sperm on VS underwear at the original trial, so that would mean that the rapist was not azoospermic and of blood group O.

KR,
Vic.

[Dupplin Muir]

Hello all

I finally got around to watching the 2002 'Horizon' on the A6 murder, and I fear that it rather fell into its usual trap of 'trust the men in the white coats'. I was also rather surprised that it used the false logic of asking 'if JH wasn't the rapist, where's the rapist's DNA?'

I believe this was mentioned on the old thread, but one counter to this argument would be that the rapist was azoospermic. The website

http://www.andrologyjournal.org/cgi/content/full/23/3/293

puts the incidence of this condition at 2%. Given some of the improbabilities and coincidences surrounding this case, odds of 50:1 seem like a racing cert!

DM (not FSS: that was me playing around with the post in Notepad and not proofreading properly!)

[Dupplin Muir]

Hello Victor

Hmm...I'm not sure I'd put too much weight on press encomia of forensic scientists: I can remember all too well the praise lavished on Professor James Cameron, before it came out that his testimony had helped to convict innocent people in both the Maxwell Confait and Lindy Chamberlain cases (as well as who knows how many others).

You might add the forensic scientists who testified against the Birmingham Six and Sally Clark.

The problem I have with Dr Whitaker is why, if the LCN DNA method is so reliable, does he pop up in criminal cases as far away as New Zealand? Surely there must be some other laboratory that would be closer?

It seems to me, after Googling about a bit, that elsewhere in the world LCN is seen as basically exculpatory: only FSS seem to think that it is capable of proving a suspect's guilt.

Here are another couple of interesting sites and I've extracted some points that seem relevant to the A6 case.

http://www.scientific.org/tutorials/articles/riley/riley.html

PCR-based testing often requires less DNA than RFLP testing and the DNA may be partially degraded, more so than is the case with RFLP. However, PCR still has sample size and degradation limitations that sometimes may be under-appreciated. PCR-based tests are also extremely sensitive to contaminating DNA at the crime scene and within the test laboratory. During PCR, contaminants may be amplified up to a billion times their original concentration. Contamination can influence PCR results, particularly in the absence of proper handling techniques and proper controls for contamination.

...

Prevention of false results involves the use of carefully applied controls and techniques. As described later, such controls and techniques can rarely guarantee that contamination hasn't influenced the results. In forensic DNA testing, some of the scientifically worst-case scenarios can be prevented by keeping DNA samples from known individuals well out of range of other items of evidence at all stages. Most forensic DNA laboratories perform negative controls, blank samples that will often detect contaminants in the laboratory. The blanks detect contaminants by showing partial or full DNA profiles representing the contaminants. Alternatively, the blank may show no profile, consistent with, but not proving that contamination didn't occur. Unfortunately, a few forensic DNA laboratories omit their controls. A few favor the controls by using special equipment on them, or by not carrying them through the entire procedure. Such practices are hazardous, especially when an important evidentiary sample has a low amount of DNA, degraded DNA, or otherwise presents as a minimal or partial (see below) sample. In short, while PCR is a useful research tool, all applications require extreme care and vigilance.

...

It is critically important to store samples in proper containers and keep known samples well-segregated from other evidence, particularly evidentiary samples that have small amounts of DNA. Paper envelopes or wax-paper folds are unsuitable containers.

...

The laboratory should be extremely careful not to overstate the scientific value of the evidence. For example, reports that a profile occurs in 1 in a billion, randomly selected individuals greatly overstate the proven error rate of the technology since false convictions based on DNA evidence have been established. Perhaps such rare match probabilities could be reached if thoroughly independent samples produced the same results in multiple, independent, non-communicating laboratories. But, for single laboratories, extremely rare match probabilities misrepresent the scientific value of technology.

http://www.aclu.org/privacy/biotech/asset_upload_file971_29096.pdf

The potential for error in DNA testing is exacerbated by the context in which labs carry out their work. Lab technicians do not typically “blind” themselves to the government’s expected or desired outcome. Studies have revealed lab notes that indicate that analysts are familiar with facts in their cases and are aware of which results will help or hurt the prosecution. For example, one set of notes stated, “Death penalty case. Need to eliminate [other individual] as a possible suspect.” It is a well-established psychological phenomenon that people tend to see what they expect to see, particularly in ambiguous situations.

I don't think that anyone can say that these caveats don't apply to FSS, since it appears that they reject any independent oversight of their work.

Again, apologies for the length of this post!

All the best

FSS

[Victor]

Whitaker

On the A6 thread the World's End case has been mentioned, and there's a link here -> http://www.guardian.co.uk/science/2007/sep/16/dna

Taking some quotes from that article.

"Whitaker is one of the world's most highly respected DNA profilers and has helped to convict a string of killers"
"Whitaker has a careful, precise manner - perhaps a result of his need for exactitude in his work"
"Whitaker is the epitome of caution"

Just thought I'd include those to counteract other comments on this thread.

KR,
Vic.