The Trials of Amanda Knox and Raffaele Sollecito: Part 30

[Stacyhs]

http://www.internationalskeptics.com/forums/showthread.php?postid=12878629#post12878629

A useful non-technical reference on the analysis of profiles of DNA mixtures:

https://www.nist.gov/featured-stories/dna-mixtures-forensic-science-explainer

Here is a quote from the above source with an analogy on DNA mixtures:

"UNCERTAINTY #2: Whose peak is it anyway?

When analyzing a DNA mixture, the alleles from all the contributors show up on the same chart. This can make it difficult to tease apart the DNA profiles of the individual contributors. To understand why this makes things complicated, recall that after amplifying the DNA, the forensic scientist has a test tube with millions of copies of the alleles in solution. Think of that test tube as a bowl of alphabet soup.

In this bowl of soup, each letter represents a different type of allele. Our suspect is named JOHN Q SUSPECT.



We analyze the soup and find that all the letters in the suspect’s name are present. Does that mean someone named JOHN Q SUSPECT contributed to the soup?

Not necessarily. There could have been two contributors named PATRICK QUEEN and JUSTIN OHR. In that case, the soup would have all the letters needed to spell JOHN Q SUSPECT, even though no person with that name contributed to the soup."

The example provided by Numbers comes directly from NIST (National Institute for Standards and Technology):

NIST has produced several PCR-based DNA Profiling Standard Reference Materials (SRMs) for the forensic community. The primary uses of these materials are validation and calibration of currently used methods for quality assurance purposes.

But what do they know? They're only scientists, four of whom are Nobel Prize winners.

Er, best stick to law.

Er, best stick to accounting.

 

[LondonJohn]

It was the bra clasp, not strap.

Well yes (though obv the clasp is part of the strap...). And Guede's DNA was not on the clasp

(I'm guessing that Vixen meant Sollecito, not Guede)

 

[LondonJohn]

The example provided by Numbers comes directly from NIST (National Institute for Standards and Technology):



But what do they know? They're only scientists, four of whom are Nobel Prize winners.

There's really very little one can do to counter opinions that are borne of scientific illiteracy, Stacy. Especially when a body such as NIST does excellent lay analogies - which are explicitly intended to explain and reveal somewhat esoteric/complex science concepts to non-scientists - and those analogies apparently remain misunderstood (or totally lacking in understanding) as well....

 

[LondonJohn]

The chutzpah. Stacyhs sets herself up as the judge and jury, whilst I am expected to jump through her hoops, when at the end of the day she dismisses anything contrary to her heroes. Justice, which is supposed to be cold, objective and neutral is not her strong point, at least not one likely to lead to a career in the judiciary that would last five minutes.

No, Vixen. You're being asked to supply evidence to support YOUR claims.

Specifically, in this instance:

1) Evidence to support your claim that the Knox family engaged in a "$2 million PR campaign" (evidence which you claimed you possessed....)

2) Evidence to support your claim that the PR activities undertaken by the Knox family included "intensive advertising in the media".


No known evidence exists (in the public domain), as yet, to support either of these claims of yours, Vixen. And it's you who made these claims. Therefore it's entirely incumbent upon you to support these claims with (credible, reliable) evidence, or else withdraw them (I can't really believe that I and others are still having to explain this concept to you).

On the other hand, you could simply admit that both claims are lies (which, in fact, they are, aren't they), and move on to raise (yet) another mole from a different burrow.....

 

[Stacyhs]

Well yes (though obv the clasp is part of the strap...). And Guede's DNA was not on the clasp

(I'm guessing that Vixen meant Sollecito, not Guede)

For accuracy's sake, the clasp is not part of the strap. The straps are the part that go over the shoulder. The clasp we are talking about is part of the back band.

According to the Vinci report, he found some alleles that were compatible with Guede, just as with Knox. But again, it could not be reliably used to identify anyone as Vinci wrote in his conclusion.

 

[LondonJohn]

For accuracy's sake, the clasp is not part of the strap. The straps are the part that go over the shoulder. The clasp we are talking about is part of the back band.

According to the Vinci report, he found some alleles that were compatible with Guede, just as with Knox. But again, it could not be reliably used to identify anyone as Vinci wrote in his conclusion.

Oh OK, I concede my lack of familiarity with bra component part names


(And the presence of some of Guede's alleles only tends to point once again to the dangers of suspect-centric "identification". It wouldn't be surprising if one could have found some (or even more) of the Pope's alleles on that noise-compromised admixture.....)

 

[Stacyhs]

Vinci found some alleles that were compatible with Guede and Knox. Just as the letters S, E, N, and T are compatible with the names Steven, Stephen, Esteban, Nestor, Stanley, and Estancia.

 

[LondonJohn]

Vinci found some alleles that were compatible with Guede and Knox. Just as the letters S, E, N, and T are compatible with the names Steven, Stephen, Esteban, Nestor, Stanley, and Estancia.

Exactly - just like the "John Q Suspect" analogy in the NIST paper.

Of course though, there are none so blind as those who will* not see.


* As a point of accuracy, the word "will" is used here in its original sense meaning "want to", not as the expression of the future tense it has come to represent. Therefore the quote in more modern language is "there are none so blind as those who do not want to see"....

 

[Vixen]

Experiment

Aim

To examine genome sequencing.

Apparatus

- Plastacine, PlayDoh or dough made from a mix of flour water and oil. Alternatively, a paper and pencil. Pen and paper.

- Four different coloured highlighter pens.

- A knife or a pair of scissors.

Method

Roll out the plastacine (or other) into two long strips, reasonably thick.

Make a mark on one of them to denote 'SIDE A'.

Roll out and insert twenty 'rungs' to make your two strips look like a ladder.

Alternatively, draw your 'ladder' on a piece of paper.

Now using the four different colour highlighters mark the 'SIDE A' rungs different colours in any order, ensuring the full range of the four is spread along 'SIDE A' of the ladder. If using paper and pencil, simply write in the letters A,B,C or D along the side of the rungs, in any order but using all four evenly, so there is five of each.

Now repeat this process on the other side of the ladder 'SIDE B' using the same method as for 'SIDE A'.

Now carefully take one end of of your 'ladder' in one hand and the other in the other hand. Carefully turn one end clockwise whilst turning the other end anti-clockwise, just once or twice.

If on paper, cut around the ladder drawing and twist, as above.

This represents the double helix of a DNA strand. (NB: In a DNA model the 'rungs' strictly speaking are four-cornered squares with a different sugar/amino-acid on each corner, but to keep it simple, we will not concern ourselves with that, here.)

STEP 2.

Now straighten out your ladder again and keeping it flat, lay it down on a flat even surface in any direction you like, back to front, upside down, obliquely or even in a curve.

Jot down the sequence you have on 'SIDE A' and then the sequence on 'SIDE B' as per your highlighter colours (see above) or your pencil annotation on a piece of paper or laptop.

These sequences represent your pairs of alleles, with opposite sides of each rung pairing with each other. For example, RED/BLUE, GREEN/GREEN, YELLOW/BLACK, BLACK/RED etc., etc., or A/B, C/C, D/B, D/A, etc, etc. Jot down your pairs, using the same order. These are your markers.

STEP 3.

Taking a knife, or a pair of scissors, cut your double helix into three sections, horizontally, so that you have two sections with seven pairs of alleles and one with six.


Mix them around.

Jot down the paired sequences of each of your sections. This represents your STR strings.


Results

Your DNA sequencing remains in a fixed particular sequence regardless of how much you stir them around, mix them up or amplify them. The STR sections are allele fragments and can still be used in DNA identification if sufficient pairs of alleles match.

Homework So, a forensic police officer now hands you the DNA/STR sequence of a possible suspect. You are asked to calculate the probability the suspect's markers matches that found at the scene of the crime, as per your analysis, above.

What is the probability that the suspect's DNA/STR marker sequence matches your sample? (Hint: work on the basis of NULL HYPOTHESIS: the probability the sample comes from a completely random section of the population and not the crime scene. At what level will you rule out it is pure chance? Hint: this could be at the 95% level or 99% or even 99.9%. [Note as the alleles are paired the probability calculation will be based on (4 x 4) ^20, (or ^7 or ^6, depending on number of loci you are looking at.]

CONCLUSION

Your marker sequences do not change position in a random higgledy piggledy manner like an anagram. A sequence showing, say:

JOHN Q SUSPECT

Cannot possibly match a suspect with the DNA/STR sequence of

PATRICK QUEEN

or

JUSTIN OHR

DNA is not an alphabetti-spaghetti soup, as this experiment demonstrates.

It will be readily seen that the probability that a random man in the street has the same sequence of markers as DNA found at a crime scene to be objectively vanishingly remote and resistant to manipulation.


Here endeth today's lesson.

 

[Welshman]

Experiment

Aim

To examine genome sequencing.

Apparatus

- Plastacine, PlayDoh or dough made from a mix of flour water and oil. Alternatively, a paper and pencil. Pen and paper.

- Four different coloured highlighter pens.

- A knife or a pair of scissors.

Method

Roll out the plastacine (or other) into two long strips, reasonably thick.

Make a mark on one of them to denote 'SIDE A'.

Roll out and insert twenty 'rungs' to make your two strips look like a ladder.

Alternatively, draw your 'ladder' on a piece of paper.

Now using the four different colour highlighters mark the 'SIDE A' rungs different colours in any order, ensuring the full range of the four is spread along 'SIDE A' of the ladder. If using paper and pencil, simply write in the letters A,B,C or D along the side of the rungs, in any order but using all four evenly, so there is five of each.

Now repeat this process on the other side of the ladder 'SIDE B' using the same method as for 'SIDE A'.

Now carefully take one end of of your 'ladder' in one hand and the other in the other hand. Carefully turn one end clockwise whilst turning the other end anti-clockwise, just once or twice.

If on paper, cut around the ladder drawing and twist, as above.

This represents the double helix of a DNA strand. (NB: In a DNA model the 'rungs' strictly speaking are four-cornered squares with a different sugar/amino-acid on each corner, but to keep it simple, we will not concern ourselves with that, here.)

STEP 2.

Now straighten out your ladder again and keeping it flat, lay it down on a flat even surface in any direction you like, back to front, upside down, obliquely or even in a curve.

Jot down the sequence you have on 'SIDE A' and then the sequence on 'SIDE B' as per your highlighter colours (see above) or your pencil annotation on a piece of paper or laptop.

These sequences represent your pairs of alleles, with opposite sides of each rung pairing with each other. For example, RED/BLUE, GREEN/GREEN, YELLOW/BLACK, BLACK/RED etc., etc., or A/B, C/C, D/B, D/A, etc, etc. Jot down your pairs, using the same order. These are your markers.

STEP 3.

Taking a knife, or a pair of scissors, cut your double helix into three sections, horizontally, so that you have two sections with seven pairs of alleles and one with six.


Mix them around.

Jot down the paired sequences of each of your sections. This represents your STR strings.


Results

Your DNA sequencing remains in a fixed particular sequence regardless of how much you stir them around, mix them up or amplify them. The STR sections are allele fragments and can still be used in DNA identification if sufficient pairs of alleles match.

Homework So, a forensic police officer now hands you the DNA/STR sequence of a possible suspect. You are asked to calculate the probability the suspect's markers matches that found at the scene of the crime, as per your analysis, above.

What is the probability that the suspect's DNA/STR marker sequence matches your sample? (Hint: work on the basis of NULL HYPOTHESIS: the probability the sample comes from a completely random section of the population and not the crime scene. At what level will you rule out it is pure chance? Hint: this could be at the 95% level or 99% or even 99.9%. [Note as the alleles are paired the probability calculation will be based on (4 x 4) ^20, (or ^7 or ^6, depending on number of loci you are looking at.]

CONCLUSION

Your marker sequences do not change position in a random higgledy piggledy manner like an anagram. A sequence showing, say:

JOHN Q SUSPECT

Cannot possibly match a suspect with the DNA/STR sequence of

PATRICK QUEEN

or

JUSTIN OHR

DNA is not an alphabetti-spaghetti soup, as this experiment demonstrates.

It will be readily seen that the probability that a random man in the street has the same sequence of markers as DNA found at a crime scene to be objectively vanishingly remote and resistant to manipulation.


Here endeth today's lesson.

It is hilarious that Vixen who has consistently shown scientific illiteracy and ignorance in her posts still feels she is in a position to give science lessons.

 

[whoanellie]

[(NB: In a DNA model the 'rungs' strictly speaking are four-cornered squares with a different sugar/amino-acid on each corner, but to keep it simple, we will not concern ourselves with that, here.)

You still haven't figured out the difference between DNA and protein? This has been going on for years now Vixen.

 

[LondonJohn]

You still haven't figured out the difference between DNA and protein? This has been going on for years now Vixen.

"A different sugar on each corner" LMAO

Pitiful

 

[whoanellie]

Your marker sequences do not change position in a random higgledy piggledy manner like an anagram. A sequence showing, say:

JOHN Q SUSPECT

Cannot possibly match a suspect with the DNA/STR sequence of

PATRICK QUEEN

or

JUSTIN OHR

DNA is not an alphabetti-spaghetti soup, as this experiment demonstrates.

It will be readily seen that the probability that a random man in the street has the same sequence of markers as DNA found at a crime scene to be objectively vanishingly remote and resistant to manipulation.

Forensic DNA profiling is not the same as DNA sequencing. Thus beginneth your error.

 

[Rolfe]

(NB: In a DNA model the 'rungs' strictly speaking are four-cornered squares with a different sugar/amino-acid on each corner, but to keep it simple, we will not concern ourselves with that, here.)

That's right up there with males are YY and males don't have mitochondrial DNA.

Vixen, you would not get a job even as a demonstrator in my lab. Even when we use simplistic models to get a point across it is required that both students and staff actually understand the basics of the biology involved.


ETA: Oops, sorry, everybody else seems to have noticed this too.

 

[Stacyhs]

Where did you get this 'experiment', Vixen? You didn't cite it and I can't find it anywhere.

Is this another made up experiment similar to your "ketchup transfer" experiment?

 

[Planigale]

Vixen still does not understand the difference between base pairs, loci and alleles. It is as if she does not understand that some of us have actually done this stuff as undergraduates, now I am not going to claim to be an expert, it was 15 years ago, but I have extracted DNA, done PCR, sequenced a small part of a gene. tagged a house keeping gene with a fluorescent marker inserted it in to a plasmid injected it into the girlfriend and now she glows in the dark when she hoovers.

ETA some of the above I will deny if any of you report me to the ethics board.

 

[TruthCalls]

Vixen, the following table represents Amanda's DNA profile and how it maps to the DNA extraction performed on sample 165B - the bra clasp.

There are three columns (A, B & C), which represent three different interpretations of the egram from the sample. Col A is what is reported in the RTIGF. Col B is every peak, with no minimum RFU value and Col C is all peaks with an RFU of 50 and above.

For each loci, I compare Amanda's profile to those of the sample as cited in each of the three interpretations. If there is a match, then I highlight the box with either yellow (found in Meredith's profile), red (Sollecito), purple (Guede) or white text on black (unique to Amanda).

As you can see, based on what was reported in the RTIGF, there were no unique allleles from Amanda. Where a minimum of 50 RFU is enforced from the egram (Col C), there are three alleles unique to Amanda. As 50 RFU is the limit Stefanoni adheres to, this is the most incriminating result one can reasonably argue. And I'm sure even Balding would tell you (as he already has) this is NOT evidence of Amanda's profile on the clasp.

In case you're confused on how to interpret this...

Consider locus D8S1179 (Amanda profile: 11, 12)

The RTIGF reports 13, 15 & 16 so no alleles match Amanda's profile and so under Col A for this loci the two boxes are not highlighted.

Per the egram, 12,13,14,15 & 16 are all above 50 RFU and so under Col C for this loci the box with 12 is highlighted (allele 12 is in Amanda's profile and this interpretation of the egram), 11 is not.

If either 11 or 12 was found in Meredith, Raffaele or Guede's profile then those boxes would be highlighted with the correct color. In this case, 12 is not in anyone else's profile so it's unique to Amanda and is highlighted as white on black.

There is far more information to consider than just this. For example, in the Vinci report, regarding this particular locus he writes;

"D8S1179: a profile 13, 15, 16 was attributed to the trace. Alleles 12 and 14 (RFU> 50) were omitted. The 12 allele having an area of 689 (less than 15% of the area of ​​the neighboring main peak (6239) could have been considered a stutter of the main allele 13 (the percentage of stutter for these small ones varies from 3 to 8% as reported by the manufacturer). This eventuality can not excluded but, in the presence of no more DNA present in the mixture at a concentration of about one tenth compared to the component greater, a stutter band is absolutely indistinguishable from a true allele that should have been attributed. For the 14 allele of the assessment (469 ie at least 45% of the peak area 15 next , (1171)) shows that this peak is not considered in the analysis has a higher percentage of 15% (if we consider the definition stutter band) and therefore must be considered in all respects effects a definable allele and attributable in the mixture."

So even though in the more liberal interpretation of the egram we see allele 12 appear, based on it's relative low RFU, it's reasonable to consider this stutter.

I can't wait to see The Machine's presentation. That should really push his creative writing skills to the limit.

 

[Stacyhs]

Thank you, TC.

 

[Stacyhs]

Vecchiotti testified that the DNA on the bra was so mixed that she could find her own loci in that mix:

1.20 pm - In the mixed track of DNA found on the hook of the bra worn by Meredith Kercher, the genetic profile of one of the experts of the Court of Appeal of Perugia can also be identified. "There's also me", explained Professor Carla Vecchiotti, answering a question from Raffaele Sollecito's defense. "Nine loci correspond to me," said the expert. Experts have argued that the trace on the hook is composed of the DNA of several subjects. One is "certainly" - it was explained - that of the student. For others - according to the experts - the genetic profiles of more people can be reconstructed. Like Sollecito but also - it has emerged today - by Professor Vecchiotti herself.

http://www.umbrialeft.it/notizie/processo-meredith-si-ritorna-aula-processo-dappello

If I'm reading the chart above correctly, Amanda matched only 5 loci whereas Vecchiotti matched 9 loci. Crime solved: Vecchiotti did it!

 

[LondonJohn]

Vixen, the following table represents Amanda's DNA profile and how it maps to the DNA extraction performed on sample 165B - the bra clasp.

There are three columns (A, B & C), which represent three different interpretations of the egram from the sample. Col A is what is reported in the RTIGF. Col B is every peak, with no minimum RFU value and Col C is all peaks with an RFU of 50 and above.

For each loci, I compare Amanda's profile to those of the sample as cited in each of the three interpretations. If there is a match, then I highlight the box with either yellow (found in Meredith's profile), red (Sollecito), purple (Guede) or white text on black (unique to Amanda).

As you can see, based on what was reported in the RTIGF, there were no unique allleles from Amanda. Where a minimum of 50 RFU is enforced from the egram (Col C), there are three alleles unique to Amanda. As 50 RFU is the limit Stefanoni adheres to, this is the most incriminating result one can reasonably argue. And I'm sure even Balding would tell you (as he already has) this is NOT evidence of Amanda's profile on the clasp.

[qimg]http://www.internationalskeptics.com/forums/picture.php?albumid=1320&pictureid=12295[/qimg]

In case you're confused on how to interpret this...

Consider locus D8S1179 (Amanda profile: 11, 12)

The RTIGF reports 13, 15 & 16 so no alleles match Amanda's profile and so under Col A for this loci the two boxes are not highlighted.

Per the egram, 12,13,14,15 & 16 are all above 50 RFU and so under Col C for this loci the box with 12 is highlighted (allele 12 is in Amanda's profile and this interpretation of the egram), 11 is not.

If either 11 or 12 was found in Meredith, Raffaele or Guede's profile then those boxes would be highlighted with the correct color. In this case, 12 is not in anyone else's profile so it's unique to Amanda and is highlighted as white on black.

There is far more information to consider than just this. For example, in the Vinci report, regarding this particular locus he writes;

"D8S1179: a profile 13, 15, 16 was attributed to the trace. Alleles 12 and 14 (RFU> 50) were omitted. The 12 allele having an area of 689 (less than 15% of the area of ​​the neighboring main peak (6239) could have been considered a stutter of the main allele 13 (the percentage of stutter for these small ones varies from 3 to 8% as reported by the manufacturer). This eventuality can not excluded but, in the presence of no more DNA present in the mixture at a concentration of about one tenth compared to the component greater, a stutter band is absolutely indistinguishable from a true allele that should have been attributed. For the 14 allele of the assessment (469 ie at least 45% of the peak area 15 next , (1171)) shows that this peak is not considered in the analysis has a higher percentage of 15% (if we consider the definition stutter band) and therefore must be considered in all respects effects a definable allele and attributable in the mixture."

So even though in the more liberal interpretation of the egram we see allele 12 appear, based on it's relative low RFU, it's reasonable to consider this stutter.

I can't wait to see The Machine's presentation. That should really push his creative writing skills to the limit.

Exactly.

And it's not just those alleles that (according to one's interpretation* of the eGram) might be present at any given locus - it's also those which are not present.

For example, at D7S820, if Knox's DNA were indeed present on the sample we would expect to see a peak at allele 9. But no such peak is present. This in itself is good evidence for the absence of Knox's DNA on the sample. And there are several other similar instances from that chart.


* Uninformed/uneducated pro-guilt commentators also seem to believe that DNA analysis is a "black and white" matter, where there's no room for subjectivity or operator bias. But of course that's nonsense, especially when one is analysing a) samples near or below the low-template range, and/or b) samples which contain admixtures of DNA from several contributors. In the case of the bra clasp, both these things are true. In these circumstances, it's vital that analysts guard strongly against subject-centric identifications, and it's equally vital that analysts are rigorous in their interpretation of true peaks from stutter and other noise. Not-a-real-doctor Stefanoni did neither......