Continuation Part Eight: Discussion of the Amanda Knox/Raffaele Sollecito case

[acbytesla]

acbytesla,

Both processes use fluorescent dyes, but they do so in very different ways. Quantitating DNA by the real-time PCR method uses the polymerase chain reaction (PCR) to increase the amount of DNA in a repetitive manner (mitosis is not an accurate description of the process). The Qubit fluorometer does not use PCR at all, but there is a dye that is specific for binding to double stranded DNA. If I have time this evening, I will come back to this subject.

Thanks Chris. Mitosis is where cells split and double, right? But PCR doesn't double the number of cells in the same way with each cycle? It just doubles the number of DNA strands without cellular division? I'm really curious how the Qubit works. I'm perusing the manual now..but the manual seems to be for the Qubit 2.0. I'm guessing that Stefanoni was using the 1.0 version of this machine? So I'm wondering if the difference between the two is significantly different?

 

[acbytesla]

Essentially yes, the difference being with Real Time that the sample is amplified (making it easier to count) and then they figure backwards from there to determine how much they started with.

So, how does Stefanoni get a too low reading time and time again? And what is she doing between each test?

 

[Kaosium]

So, how does Stefanoni get a too low reading time and time again? And what is she doing between each test?

She got the "too lows" because the samples she was testing didn't have any DNA in them or not enough for the Qubit to count. Those 'too lows' weren't for the same sample, they were the results of her testing all the areas on the knife (36B 36C etc) where the only one showing positive for DNA was 36A, the knife handle with Amanda's DNA on it from cutting food, all the rest were measured as 'too low' for the Qubit to register.

There's no amplification of DNA involved with the Qubit, essentially it just 'marks' it with a dye and then they measure that to get a result. What's suspicious is how she decided 36B was positive for amplification but not 36C (or any of the other 'too lows' ) despite getting the same result for all those samples. It suggests she generated that paperwork after the fact to allow for her to have a reason to have proceeded to amplification with 36B but at the same time not disclose any of the other records of the samples she got 'too lows' on as they were not considered 'positive for amplification.'


It also shows Stefanoni was lying when she claimed she'd quantified 36B with Real Time DNA and gotten a result of 'hundreds of picograms.' That's basically what Conti and Vecchiotti were getting at here:

Regarding the interpretation of the quantification, it should be noted that on page 78, the following is stated: “the samples testing positive to quantification (samples A and B) were subjected to amplification and subsequent capillary electrophoresis…”.

As regards the quantification of sample A (knife handle) the results obtained with the Qubit Fluorimeter™ show that the concentration of DNA in this sample was equal to 0.08 ng/μl. Taking into account that the “quantity of extract” was 50 μl (c.f. SAL), and multiplying 0.08 ng/μl x 50 μl, the total [amount] of DNA was equal to 4 ng: certainly a significant quantity, which allowed sample A to be considered positive to quantification.

On the other hand, it is not possible to comprehend the criteria adopted in the assessment of the positive quantification result for sample B and the negative result for sample C, given that the same result, “too low”, was obtained for both samples: that is, a value which must be considered not only below the sensitivity threshold of the Fluorimeter indicated by the manual (DNA concentrations equal to 0.2 ng/μl), but below 0.08 ng/μl, a value which the Fluorimeter detected for sample A.

Nor is it comprehensible, considering the negative results on sample B, what Dr. Stefanoni reported during the GUP questioning (page 178) where she stated that the DNA in sample B, quantified with Real Time PCR (it is recalled that such quantification as confirmed during the hearing was never carried out or, at least, no documentation was provided to support this claim), was “in the order of some hundreds of picograms”, a value which does not appear in any of the documents provided to us (SAL, Fluorimeter report, Real Time report, RTIGF).

emphasis retained

 

[Kaosium]

So, how does Stefanoni get a too low reading time and time again? And what is she doing between each test?

Incidentally, she did do something else, that is 'concentrate' the sample 36B by reducing the amount of liquid relative to the DNA by using the Speed Vac machine. How she 'knew' to keep concentrating that sample and none of the others despite the fact she got 'too low' from the Qubit (which is the same as you'd get with nothing at all) for all of them is another one of the mysteries of this case with ponderables unpalatable to Stefanoni.

ETA, she did indeed concentrate samples 36C-G, as is recorded in C&V here in the first paragraph.

“The amplification of the autosomal STRs was performed according to the methods already reported on page 31; the samples which tested positive to quantification (samples A and B) were subjected to amplification and subsequent capillary electrophoresis. The samples which tested negative to quantification (samples C, D, E, F, G) were analyzed following concentration using the Speed-Vac SC110 instrument, ‘Savant’ brand”.

 

[acbytesla]

She got the "too lows" because the samples she was testing didn't have any DNA in them or not enough for the Qubit to count. Those 'too lows' weren't for the same sample, they were the results of her testing all the areas on the knife (36B 36C etc) where the only one showing positive for DNA was 36A, the knife handle with Amanda's DNA on it from cutting food, all the rest were measured as 'too low' for the Qubit to register.

There's no amplification of DNA involved with the Qubit, essentially it just 'marks' it with a dye and then they measure that to get a result. What's suspicious is how she decided 36B was positive for amplification but not 36C (or any of the other 'too lows' ) despite getting the same result for all those samples. It suggests she generated that paperwork after the fact to allow for her to have a reason to have proceeded to amplification with 36B but at the same time not disclose any of the other records of the samples she got 'too lows' on as they were not considered 'positive for amplification.'


It also shows Stefanoni was lying when she claimed she'd quantified 36B with Real Time DNA and gotten a result of 'hundreds of picograms.' That's basically what Conti and Vecchiotti were getting at here:

I am confused Kaosium. For some reason Stefanoni chose to use the Qubit on some select samples to determine if there was DNA? With the Real time PCR machine. All samples undergo the PCR process to see if there is DNA and when DNA is detected, a tech then subjects the sample to electropherisis.

The Real time machine can combine DNA detection and PCR in the same process while the Qubit Flourimeter is for detection only? Normally only a sample with positive DNA result would be subjected to PCR and then electropherisis. But in this case, multiple locations on the the knife (36) were not subjected to the Real time machine for DNA detection but were tested by the Qubit Flourimeter instead and one virtually all of these samples (not just 36B) Stefanoni got a "too low" result but selectively decided to put 36B through PCR using the Real time machine?

 

[Desert Fox]

Kaosium
Thanks, I understood most of that

 

[Kaosium]

I am confused Kaosium. For some reason Stefanoni chose to use the Qubit on some select samples to determine if there was DNA? With the Real time PCR machine. All samples undergo the PCR process to see if there is DNA and when DNA is detected, a tech then subjects the sample to electropherisis.

The Real time machine can combine DNA detection and PCR in the same process while the Qubit Flourimeter is for detection only? Normally only a sample with positive DNA result would be subjected to PCR and then electropherisis. But in this case, multiple locations on the the knife (36) were not subjected to the Real time machine for DNA detection but were tested by the Qubit Flourimeter instead and one virtually all of these samples (not just 36B) Stefanoni got a "too low" result but selectively decided to put 36B through PCR using the Real time machine?

It doesn't sound to me like you're very confused, you seem to have picked this up pretty quickly. Unless it's a matter of being confused by why Stefanoni did what she did, in which case there's quite a crowd forming which includes both Conti and Vecchiotti as well as myself. I always liked the way they put 'it is impossible to comprehend the criteria' regarding this issue.

As for using the Qubit as opposed to simply running Real-Time PCR there was an issue with the Real Time Machine--at least for that sample. For samples D-E-F-G, all negative, she did run them through real time PCR, however as I understand that it was a different day, significantly later as I recall off the top of my head. So in other words she failed to run Real-Time PCR on 'B' (and C too incidentally) which she considered 'positive' for amplification, (despite getting the same 'too low') yet did run Real Time PCR for the samples which were negative which is damned strange. You read that right, and no it doesn't make sense.


Here's that section in C&V, I'd quote the part but I'm getting sick of having to add all the emphasis tags!

Note also that I did recently read some sort of confirmation on the Real Time Machine being down, it was in one of Diocletus' posts not that long ago. She'd had a contamination incident and stopped using that machine for a while. Another very interesting thing about the knife is that the date of the amplification is unknown for the relevant (A&B) knife samples as I understand it.

 

[Diocletus]

What's suspicious is how she decided 36B was positive for amplification but not 36C (or any of the other 'too lows' ) despite getting the same result for all those samples. It suggests she generated that paperwork after the fact to allow for her to have a reason to have proceeded to amplification with 36B but at the same time not disclose any of the other records of the samples she got 'too lows' on as they were not considered 'positive for amplification.'

People tend to get this mixed up. She submitted all of the Qubit samples to amplification. Thus, the irregularity is not that 36b was submitted to amplification, rather, it is that we do not have an egram for 36c, which was also amplified. For that matter, we are missing the egrams for 90% of the Qubit samples--and those are what she is trying to hide by making a distinction between positive/negative quantification, when in fact she didn't really perform any kind of effective quantification at all.

 

[Diocletus]

Note also that I did recently read some sort of confirmation on the Real Time Machine being down, it was in one of Diocletus' posts not that long ago. She'd had a contamination incident and stopped using that machine for a while. Another very interesting thing about the knife is that the date of the amplification is unknown for the relevant (A&B) knife samples as I understand it.

I thought that I had read sometime that she gave some explanation for why she started to use the Qubit, but I would like a cite for that if anyone has it. What I do know is this:

There were two (and only two) batches of Real Time that show a baseline increase of about 30%. After the first of these batches she suddenly switched to the Qubit. After the second of these batches, she abandoned use of the Real Time machine and only thereafter used the Fast Time machine. This second batch is the one that shows a contaminated positive control.

 

[Strozzi]

. . . Note also that I did recently read some sort of confirmation on the Real Time Machine being down, it was in one of Diocletus' posts not that long ago. She'd had a contamination incident and stopped using that machine for a while. Another very interesting thing about the knife is that the date of the amplification is unknown for the relevant (A&B) knife samples as I understand it.

Would Stefanoni have known it was contamination? I want to understand if she is just dumb and missed it, or if she knew it was contamination and knowingly lied when she testified in court that there had never been an incident of contamination in her (DNA) lab.

 

[Kaosium]

People tend to get this mixed up. She submitted all of the Qubit samples to amplification. Thus, the irregularity is not that 36b was submitted to amplification, rather, it is that we do not have an egram for 36c, which was also amplified. For that matter, we are missing the egrams for 90% of the Qubit samples--and those are what she is trying to hide by making a distinction between positive/negative quantification, when in fact she didn't really perform any kind of effective quantification at all.

Yes, as I noted in my next post, though oddly enough those others got Real Time PCR! What I meant regarding her paperwork was where she marked 'B' as being positive for amplification but didn't do that for 'C' as per this record where A and B get checkmarks but all the rest get an 'X.'

How long after the knife was announced to the world (which off the top of my head was about November 14th, 2007) were the other samples (D-F) run?

 

[acbytesla]

It doesn't sound to me like you're very confused, you seem to have picked this up pretty quickly. Unless it's a matter of being confused by why Stefanoni did what she did, in which case there's quite a crowd forming which includes both Conti and Vecchiotti as well as myself. I always liked the way they put 'it is impossible to comprehend the criteria' regarding this issue.

As for using the Qubit as opposed to simply running Real-Time PCR there was an issue with the Real Time Machine--at least for that sample. For samples D-E-F-G, all negative, she did run them through real time PCR, however as I understand that it was a different day, significantly later as I recall off the top of my head. So in other words she failed to run Real-Time PCR on 'B' (and C too incidentally) which she considered 'positive' for amplification, (despite getting the same 'too low') yet did run Real Time PCR for the samples which were negative which is damned strange. You read that right, and no it doesn't make sense.


Here's that section in C&V, I'd quote the part but I'm getting sick of having to add all the emphasis tags!

Note also that I did recently read some sort of confirmation on the Real Time Machine being down, it was in one of Diocletus' posts not that long ago. She'd had a contamination incident and stopped using that machine for a while. Another very interesting thing about the knife is that the date of the amplification is unknown for the relevant (A&B) knife samples as I understand it.

Unfortunately, I have to read this stuff over and over again, because it seems hard to follow.

I think what is confusing is that Stefanoni seems to use at minimum 3 different methods for achieving her results.

The Real Time machine seems great for addressing many samples at once and combining both DNA detection and PCR amplification if desired. But it also seems capable of performing these tests indivdiually and seperately. However one is putting a tray or plate into the machine capable of testing even hundreds of samples at once. (24 to 384 well plate) I'm confused how the Speed VAC comes into play and how it is used normally and how and why it was used here.

The Qubit seems designed more for testing single samples one at a time? For some reason it seems as if the Qubit is just performing spectroscopy on a given sample? ..but maybe I'm wrong about that. So now reviewing this, I'm guessing that Stefanoni tested all of the knife samples (36) with the Qubit instead of the Real time machine. That she only got a positive result from the Sample A (4 ng) and got a negative result on Samples B through G...how she ended up with 10 "too low is confusing to me as that is only 6 samples..what were the other 4?) That Stefanoni put the entire plate into the Real Time 7700 and got positive DNA results from 36A and 36B and then ran those through electropheris?

Or did I miss something? Or did I miss a lot?

 

[Kaosium]

I thought that I had read sometime that she gave some explanation for why she started to use the Qubit, but I would like a cite for that if anyone has it. What I do know is this:

I have the same 'I thought I had read' memory regarding using the Qubit but unfortunately I don't recall where I read it and would also like a cite if anyone knows where she addressed the issue.

There were two (and only two) batches of Real Time that show a baseline increase of about 30%. After the first of these batches she suddenly switched to the Qubit. After the second of these batches, she abandoned use of the Real Time machine and only thereafter used the Fast Time machine. This second batch is the one that shows a contaminated positive control.

Thanks for chiming in on that.

 

[Strozzi]

Unfortunately, I have to read this stuff over and over again, because it seems hard to follow.

I think what is confusing is that Stefanoni seems to use at minimum 3 different methods for achieving her results.

The Real Time machine seems great for addressing many samples at once and combining both DNA detection and PCR amplification if desired. But it also seems capable of performing these tests indivdiually and seperately. However one is putting a tray or plate into the machine capable of testing even hundreds of samples at once. (24 to 384 well plate) I'm confused how the Speed VAC comes into play and how it is used normally and how and why it was used here.

The Qubit seems designed more for testing single samples one at a time? For some reason it seems as if the Qubit is just performing spectroscopy on a given sample? ..but maybe I'm wrong about that. So now reviewing this, I'm guessing that Stefanoni tested all of the knife samples (36) with the Qubit instead of the Real time machine. That she only got a positive result from the Sample A (4 ng) and got a negative result on Samples B through G...how she ended up with 10 "too low is confusing to me as that is only 6 samples..what were the other 4?) That Stefanoni put the entire plate into the Real Time 7700 and got positive DNA results from 36A and 36B and then ran those through electropheris?

Or did I miss something? Or did I miss a lot?

My needs are simpler than acbytesla's. I just want to know which machine manufacturers Amanda should sue in a US court?

 

[Kaosium]

My needs are simpler than acbytesla's. I just want to know which machine manufacturers Amanda should sue in a US court?

I almost chimed in on this when you brought it up, but you're out of luck. Those machines come with disclaimers and in fact one of them (or it may have been the Genemapper software) says that it's not recommended to attempt analyzing anything below 150 RFUs! There isn't a single allele on the knife that gets anywhere close to 150 RFUs, most are between 20 and 40 RFUs with a few over fifty. Off the top of my head that wipes out most of Raffaele's profile on the clasp as well, I believe, at least for the autosomal one.

 

[Strozzi]

I almost chimed in on this when you brought it up, but you're out of luck. Those machines come with disclaimers and in fact one of them (or it may have been the Genemapper software) says that it's not recommended to attempt analyzing anything below 150 RFUs! There isn't a single allele on the knife that gets anywhere close to 150 RFUs, most are between 20 and 40 RFUs with a few over fifty. Off the top of my head that wipes out most of Raffaele's profile on the clasp as well, I believe, at least for the autosomal one.

Then why does the machine's printout print results in a normal format, instead of printing WARNING, THESE RESULTS MAY BE FALSE. GARBAGE IN, GARBAGE OUT! . These machines are sold to police forensic labs and it is foreseeable that the printouts will be used to pull the woll over the eyes of ordinary people. Why is the disclaimer not printed in the data output? It is important enough to include in the operating manuals; why not in the printout?

 

[Desert Fox]

My needs are simpler than acbytesla's. I just want to know which machine manufacturers Amanda should sue in a US court?

When somebody use every tool in the sled to see which she can abuse the most, can you blame the equipment?

 

[Kaosium]

Unfortunately, I have to read this stuff over and over again, because it seems hard to follow.

I think what is confusing is that Stefanoni seems to use at minimum 3 different methods for achieving her results.

The Real Time machine seems great for addressing many samples at once and combining both DNA detection and PCR amplification if desired. But it also seems capable of performing these tests indivdiually and seperately. However one is putting a tray or plate into the machine capable of testing even hundreds of samples at once. (24 to 384 well plate) I'm confused how the Speed VAC comes into play and how it is used normally and how and why it was used here.

The Qubit seems designed more for testing single samples one at a time? For some reason it seems as if the Qubit is just performing spectroscopy on a given sample? ..but maybe I'm wrong about that. So now reviewing this, I'm guessing that Stefanoni tested all of the knife samples (36) with the Qubit instead of the Real time machine. That she only got a positive result from the Sample A (4 ng) and got a negative result on Samples B through G...how she ended up with 10 "too low is confusing to me as that is only 6 samples..what were the other 4?) That Stefanoni put the entire plate into the Real Time 7700 and got positive DNA results from 36A and 36B and then ran those through electropheris?

Or did I miss something? Or did I miss a lot?

As for the 'too lows' being more than six, she didn't just use the Qubit on the knife as I understand, I'd need to see the rest of the page that Randy (W--surfer) recently posted from Frank Sfarzo to see what the rest of them are for.

The Qubit is not at all suited for low template work, it's detection level doesn't go low enough. Remember that with many DNA samples there will be plenty (relatively) of material to take a sample of. Someone who has bleed out might well have liters of blood on the floor with gobs and gobs (relatively) of DNA in it. Thus one takes a small bit of that and gets a sample and puts part of that through the Qubit and gets a value per microliter that amounts to having 10 nanograms (total) in the tube. As most of those machines are designed for samples of about one nanogram you'd then split your sample into tenths and have ten one nanogram sized samples to amplify. That's one of the main reasons for the quantification step, most of the machines and software are designed for samples of 0.5-1.5 nanograms. Too much and it may register off the chart, too little and there may be difficulty interpreting the results.

 

[acbytesla]

Then why does the machine's printout print results in a normal format, instead of printing WARNING, THESE RESULTS MAY BE FALSE. GARBAGE IN, GARBAGE OUT! . These machines are sold to police forensic labs and it is foreseeable that the printouts will be used to pull the woll over the eyes of ordinary people. Why is the disclaimer not printed in the data output? It is important enough to include in the operating manuals; why not in the printout?

The problem to a certain degree Strozzi is that it takes multiple machines to get a result. For example. Stefanoni tests 36B for DNA gets a negative result with the Qubit then subjects 36B to the Speedvac to concentrate the DNA and then runs it through electropherisis. It's not the manufacturer's fault that the tech misused the equipment and the Genemapper software has no idea if there was "garbage in".

 

[Chris_Halkides]

how not to keep a laboratory notebook

IIRC in the "Math on Trial" thread, Leila said that Stefanoni marked the quantitation of 36B as positive long after the quantitation experiment, because she got an egram. If this is an accurate summary, then it is an object lesson in how not to keep laboratory records. You are supposed to write an up-to-the-minute summary in your lab notebook, so that later thoughts do not color your observations.