Shall we try to understand why the LCNDNA in the Scenic is inconclusive

[debunker]

What is LCNDNA analysis  and how does it differ from ordinary DNA analysis?

Ordinary DNA analysis has a considerable amount of DNA available for analysis. The sample is acted on chemically to separate the DNA into smaller strands and then identification of strands is made looking for the sites required and determining which markers are attached to those sites. What results is a list of sites and markers allowing clear identification to be made.

LCNDNA analysis is a technique developed to determine what information is available in an inadequate sample for full DNA analysis. It works by causing the very small amount of DNA information to produce copies of itself until enough DNA is produced to start drawing conclusions from it. There is not as much information available as in normal DNA testing and site/marker pairing may be much more difficult. It is also open to major problems with cross contamination as the amounts being worked on are very close to those amounts transferred naturally to a sample by simple contact if the sample is not placed immediately in a safegiarded environment to maintain the chain of evidence. Even breathing on the sample, or a tiny particle of skin detritus could result in cross contamination.

[debunker]

So what about the back of the scenic. We know that the car had been used by Gerry, Kate, Sean and Amelie for three weeks and that rubbish had been carried in the back of the car. It is almost inconceivable that some of their DNA had not been deposited in various parts of the car including the wheel well. A mixture of DNA from any or all of those four people would show up under LCNDNA analysis as potentially from Madeleine, even if she had never been in that car.

[Benice]

What is LCNDNA analysis  and how does it differ from ordinary DNA analysis?

Ordinary DNA analysis has a considerable amount of DNA available for analysis. The sample is acted on chemically to separate the DNA into smaller strands and then identification of strands is made looking for the sites required and determining which markers are attached to those sites. What results is a list of sites and markers allowing clear identification to be made.

LCNDNA analysis is a technique developed to determine what information is available in an inadequate sample for full DNA analysis. It works by causing the very small amount of DNA information to produce copies of itself until enough DNA is produced to start drawing conclusions from it. There is not as much information available as in normal DNA testing and site/marker pairing may be much more difficult. It is also open to major problems with cross contamination as the amounts being worked on are very close to those amounts transferred naturally to a sample by simple contact if the sample is not placed immediately in a safegiarded environment to maintain the chain of evidence. Even breathing on the sample, or a tiny particle of skin detritus could result in cross contamination.

Just to clarify Db.  Could this cross contamination (e.g.from breathing on it etc) have already happened before the sample reached the lab - or are you just referring to what could happen during the test procedures?

[debunker]

What was discovered by the FSS?

The sample was too small for regular DNA analysis so was submitted to LCNDNA analysis using replication of the strands. The result was a collection of (I think) 39 markers/locations meaning that at least two people were involved; the FSS with access to the full information estimated that at least three and maybe five people contributed to the sample.

In this set of markers, 15 were the same as possessed by Maddie (not surprising as there is a strong possibility that some of the DNA traces may have come from people using the car (including the McCannfamily who between them would have all of Madeleine's markers- 50% in each parent, 100% if both parents, 25% each sib and about 50% for both sibs.

Even ignoring family contribution, the fact that it cannot be said that all 15 of the markers indicating Madeleine came from a single person. Such marker/site signs are shared by many people across an historically interbreeding  population.

[debunker]

What is LCNDNA analysis  and how does it differ from ordinary DNA analysis?

Ordinary DNA analysis has a considerable amount of DNA available for analysis. The sample is acted on chemically to separate the DNA into smaller strands and then identification of strands is made looking for the sites required and determining which markers are attached to those sites. What results is a list of sites and markers allowing clear identification to be made.

LCNDNA analysis is a technique developed to determine what information is available in an inadequate sample for full DNA analysis. It works by causing the very small amount of DNA information to produce copies of itself until enough DNA is produced to start drawing conclusions from it. There is not as much information available as in normal DNA testing and site/marker pairing may be much more difficult. It is also open to major problems with cross contamination as the amounts being worked on are very close to those amounts transferred naturally to a sample by simple contact if the sample is not placed immediately in a safegiarded environment to maintain the chain of evidence. Even breathing on the sample, or a tiny particle of skin detritus could result in cross contamination.

Just to clarify Db.  Could this cross contamination (e.g.from breathing on it etc) have already happened before the sample reached the lab - or are you just referring to what could happen during the test procedures?

Anytime before the replication procedure started would be dangerous.

You are probably referring to the myth used by [ censored word ]s that Lowe said that people from his lab contributed their DNA to the sample. What he actually said was that the identification was at such a low level of confidence because the 39 (was it 37?) alleles would be found in the admixture of any 3-5 western European's DNA recovered in this manner and with no data about the provenance- and in passing said that it was possible that such a sample from people in his lab could produce these alleles.

[debunker]

Why is ordinary DNA analysis so specific.

If you know that you have only 20 markers and they are all from the same person, each ID decreases the possibility of the sample being from someone else.

Suppose for arguments sake that only one person in a hundred had each marker (this is a gross simplification) and that the world population is 6 billion.

One marker reduces the potential population to 600 000 000

Two to 60 000 000

Three to 6 000 000

Four to 600 000

Five to 60 000

Six to 6000

Seven to 600

Eight to 60

Nine to 6

Ten and onwards to less than one, thus IDing a single individual to a certain level of assurance.


This power relationship does not hold for mixed samples- we are now doing what is called perming- the math is really quite difficult.

Say there were 40 markers found (for simplicity of calculation and five potential donors.

Marker one has a one chance in five of being from the target person and therefore the chance of this being from that person or from another person donating to the sample has to be calculated. This goes for each of the markers identified. So you do not get the progressive certainty that straight DNA analysis gives you as you have to allow for the marker being examined being from a different donor. In fact the certainty is so much reduced that any admixture of five different donor's partial DNA markers is that about fifty percent of them would include the fifteen markers identified.

And this excludes the possibility that one or more of the donors may have been relations with shared markers in abundance.

But don't think that this is going to convince the ............

[debunker]

Some cites:

http://www.cps.gov.uk/publications/prosecution/lcn_testing.html

http://en.wikipedia.org/wiki/Low_copy_number

http://www.promega.co.uk/resources/articles/profiles-in-dna/lcn-dna-analysis-limitations-prevent-general-acceptance/?__utma=1.492364517.1366882780.1366882780.1366882780.1&__utmb=1.1.10.1366882780&__utmc=1&__utmx=-&__utmz=1.1366882780.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none)&__utmv=-&__utmk=63540015

http://www.theforensicinstitute.com/news/low-copy-number-dna-and-the-forensic-institute.html

[Carana]

If we take a sample of tissue or blood or other substancefrom a person or a crime scene, and this is found to have complete unbroken strands of DNA in it, then ordinary DNA analysis can be done.

This is carried out by a process that strips the DNA strands and allows scientists to identify particular short lengths of DNA with a recognisable pattern of complex chemical elements attached in certain places.

It has been shown that these chosen areas (a different number and different locations in different jurisdictions) contain sufficient variety of structure to make rational decisions about whether different samples do in fact match to a level of probability suitable to make a scientifically and forensically valid conclusion.

This complexity is because these twenty sites (In current usage) have a variety of chemical structures (called here 'markers' (elsewhere alleles) which are associated with these particular locations. Each area has a series of possible markers associated with it, and some markers may attach to more than one different sites. The number of markers for each site totals between 3 and 20 or so. Some are very rare and some are very common associations, but with 20 possible areas and several hundred in total markers there is sufficient information to exclude everyone else from identification to very high levels of probability.

Different jurisdictions make different assessments of how many markers are required for such identification to be legally valid.

Anything contentious so far?

I have a slight problem DB. The distinction between "markers" and "alleles" has been a source of confusion throughout this case. My understanding is that the forensic labs look for two alleles at each locus used in their forensic analysis system. The locus (site) is often referred to as a marker.

[debunker]

If we take a sample of tissue or blood or other substancefrom a person or a crime scene, and this is found to have complete unbroken strands of DNA in it, then ordinary DNA analysis can be done.

This is carried out by a process that strips the DNA strands and allows scientists to identify particular short lengths of DNA with a recognisable pattern of complex chemical elements attached in certain places.

It has been shown that these chosen areas (a different number and different locations in different jurisdictions) contain sufficient variety of structure to make rational decisions about whether different samples do in fact match to a level of probability suitable to make a scientifically and forensically valid conclusion.

This complexity is because these twenty sites (In current usage) have a variety of chemical structures (called here 'markers' (elsewhere alleles) which are associated with these particular locations. Each area has a series of possible markers associated with it, and some markers may attach to more than one different sites. The number of markers for each site totals between 3 and 20 or so. Some are very rare and some are very common associations, but with 20 possible areas and several hundred in total markers there is sufficient information to exclude everyone else from identification to very high levels of probability.

Different jurisdictions make different assessments of how many markers are required for such identification to be legally valid.

Anything contentious so far?

I have a slight problem DB. The distinction between "markers" and "alleles" has been a source of confusion throughout this case. My understanding is that the forensic labs look for two alleles at each locus used in their forensic analysis system. The locus (site) is often referred to as a marker.

That was why I was trying to avoid complex words. Site and marker are simplifications but do not affect the calculations or truth of what is being said. Additionally Allele and Locus are complex latinate worlds with non-generalist acceptance whereas site and marker are simple words.

[Carana]

If we take a sample of tissue or blood or other substancefrom a person or a crime scene, and this is found to have complete unbroken strands of DNA in it, then ordinary DNA analysis can be done.

This is carried out by a process that strips the DNA strands and allows scientists to identify particular short lengths of DNA with a recognisable pattern of complex chemical elements attached in certain places.

It has been shown that these chosen areas (a different number and different locations in different jurisdictions) contain sufficient variety of structure to make rational decisions about whether different samples do in fact match to a level of probability suitable to make a scientifically and forensically valid conclusion.

This complexity is because these twenty sites (In current usage) have a variety of chemical structures (called here 'markers' (elsewhere alleles) which are associated with these particular locations. Each area has a series of possible markers associated with it, and some markers may attach to more than one different sites. The number of markers for each site totals between 3 and 20 or so. Some are very rare and some are very common associations, but with 20 possible areas and several hundred in total markers there is sufficient information to exclude everyone else from identification to very high levels of probability.

Different jurisdictions make different assessments of how many markers are required for such identification to be legally valid.

Anything contentious so far?

I have a slight problem DB. The distinction between "markers" and "alleles" has been a source of confusion throughout this case. My understanding is that the forensic labs look for two alleles at each locus used in their forensic analysis system. The locus (site) is often referred to as a marker.

That was why I was trying to avoid complex words. Site and marker are simplifications but do not affect the calculations or truth of what is being said. Additionally Allele and Locus are complex latinate worlds with non-generalist acceptance whereas site and marker are simple words.

I'm not sure that I've adequately explained my concern....

[debunker]

I see what you mean. If people can just think site and molecule then.

[Carana]

Where do we get our DNA from? We get roughly half from our mother and half from our father. So each child will share on average 50% of their mother's DNA and 50% of their father's DNA. This includes the pattern of positions and markers used to identify people using DNA.

Why did Madeleine have only 19 potential area/marker matches? Because Kate and Gerry shared on marker/position piece of information, meaning that only 19 different ones were available rather than 20.

Full siblings will share on average 25% of such position/marker pairs.

Any problems?

Perhaps this could be a better starting point?



NB: Forensic scientists are not examining the entire DNA of a person.

[Carana]

I see what you mean. If people can just think site and molecule then.

I'm interested in your explanation.

However, I think you're going too fast for common mortals (including myself) to follow. And those who are aware of how LCNDNA works probably already understand what the issue is.

[debunker]

The essence is that there is far more information in a clean example that is not contaminated.

LCNDNA analysis is firstly open to contamination, and if it is incomplete and mixed ( as the Scenic sample was) holds less information.

[debunker]

So I assume from the lack of any attempt at rebuttal that we all agree that there is no evidence that Madeleine's DNA was found in the Scenic.

Maybe people will cease claiming that that is true.