Cont: The Trials of Amanda Knox and Raffaele Sollecito: Part 32

[Numbers]

In the US state of Minnesota, a man who was convicted in 2001 of the 1998 murder of an elderly woman and sentenced to life in prison was released on 7 January 2026 based on a review of the case by the conviction review unit of the Minnesota Attorney General's office. Here are some relevant excerpts from a news article* (italics and bolding are my emphasis):

The [conviction review] unit detailed in a 119-page report that Mr. Pippitt’s conviction was based on unreliable testimony from two witnesses that should have never been presented to the jury. Both witnesses have since recanted their testimony.
.....
The unit’s investigators noted that the lead prosecutor “presented a case theory that conflicted with objective evidence,” and that he was disbarred in Minnesota in 2007.

[The conviction review unit] wrote [in its report] that Ms. Malin’s death “was unquestionably tragic,” leaving “a void in the community that could not be filled, even with the proper identification of the true murderer.”

“And yet, despite the desire for someone to atone for the crime, the atonement cannot be placed on just anybody. Otherwise, it is not justice that is served, it is convenience.”

I believe that the identification of Knox and Sollecito as suspects in the murder/rape of Kercher was, in large part, a matter of convenience for the key Italian authorities.

* Minnesota Man Is Freed After Serving 25 Years for Murder He Did Not Commit
The state’s conviction review unit concluded that Brian Pippitt, 63, was not involved in the 1998 murder of an 84-year-old woman, for which he was serving a life sentence.

https://www.nytimes.com/2026/01/08/us/brian-pippitt-wrongful-murder-conviction.html

 

[TomG]

I came across the link below from Chris H's blog while trying to find out more about the luminosity of the Luminol results at VDP7. It gives results for both liquid and burnt blood.

www.abacusdiagnostics.com/Detecting_Burnt_Bloodstain_Samples_with_Light_Emitting_Blood_Enhancement_Reagents.pdf

TMB doesn't feature on the chart; however, the way I see it is that if the sensitivity of Luminol is at 1:5,000,000 and TMB is at 1:1,000,000, it's only a fivefold difference, which doesn't seem to be that much. The chart shows a dilution of 1:800,000 recorded at 361 RFUs, so I would put TMB at 1:1,000,000, recording at around 289 RFUs. You have a signal of 100 RFUs not being detectable by the naked eye, so we are entering the territory of very low luminosity when the Luminol reactions at VDP7 were quite vibrant. Vibrant enough for Garofano to conclude that they must be blood.

If TMB at a sensitivity of 1:1,000,000 recorded 289 RFUs at it lowest detection rate and no visible signal is recordable at 100 RFUs then the Luminol traces at VDP7 would have to be invisible or near to invisible, IMO for TMB not to detect it.

I'm no scientist, and I've muddled my way through this, so I'm looking for a better-informed consensus in layman's terms.

Hoots

 

[Planigale]

View attachment 67883
I came across the link below from Chris H's blog while trying to find out more about the luminosity of the Luminol results at VDP7. It gives results for both liquid and burnt blood.

www.abacusdiagnostics.com/Detecting_Burnt_Bloodstain_Samples_with_Light_Emitting_Blood_Enhancement_Reagents.pdf

TMB doesn't feature on the chart; however, the way I see it is that if the sensitivity of Luminol is at 1:5,000,000 and TMB is at 1:1,000,000, it's only a fivefold difference, which doesn't seem to be that much. The chart shows a dilution of 1:800,000 recorded at 361 RFUs, so I would put TMB at 1:1,000,000, recording at around 289 RFUs. You have a signal of 100 RFUs not being detectable by the naked eye, so we are entering the territory of very low luminosity when the Luminol reactions at VDP7 were quite vibrant. Vibrant enough for Garofano to conclude that they must be blood.

If TMB at a sensitivity of 1:1,000,000 recorded 289 RFUs at it lowest detection rate and no visible signal is recordable at 100 RFUs then the Luminol traces at VDP7 would have to be invisible or near to invisible, IMO for TMB not to detect it.

I'm no scientist, and I've muddled my way through this, so I'm looking for a better-informed consensus in layman's terms.

Hoots

I did the maths and posted it here several years ago. The sensitivity of TMB for blood which contains lots of iron which is what it detects is greater than the ability of PCR tests at the time to detect DNA in blood (which blood actually contains relatively little of*). TMB would be positive with dilute blood stains which would be PCR negative. PCR positive TMB negative samples aren't blood. TMB can detect a few drops of blood diluted in a bashful of water.

*Red cells platelets and plasma contain no DNA, only white cells a very small component of blood contain DNA.

 

[TruthCalls]

I did the maths and posted it here several years ago. The sensitivity of TMB for blood which contains lots of iron which is what it detects is greater than the ability of PCR tests at the time to detect DNA in blood (which blood actually contains relatively little of*). TMB would be positive with dilute blood stains which would be PCR negative. PCR positive TMB negative samples aren't blood. TMB can detect a few drops of blood diluted in a bashful of water.

*Red cells platelets and plasma contain no DNA, only white cells a very small component of blood contain DNA.

So, in other words, 6 of 9 Luminol revealed samples in the cottage, which tested positive for DNA, can't be blood based on this analysis.

How sure are you of your conclusion that in 2007, PCR positive TMB negative samples couldn't be blood? Are you aware of any forensic experts who have documented this?

Between this, the brightness of the Luminol reaction, and that Meredith's DNA profile was not found in 28 of the 31 Luminol revealed samples all suggest these samples were not made from Meredith's blood. But I'm not sure we have certainty about this relationship (TMB/PCR) nor am I sure we have all we need to properly evaluated the brightness of the Luminol reaction and what that means. I often accuse pro-guilt of seeing what they want to see, and I would prefer to not make the same mistake myself.

 

[TruthCalls]

View attachment 67883
I came across the link below from Chris H's blog while trying to find out more about the luminosity of the Luminol results at VDP7. It gives results for both liquid and burnt blood.

www.abacusdiagnostics.com/Detecting_Burnt_Bloodstain_Samples_with_Light_Emitting_Blood_Enhancement_Reagents.pdf

TMB doesn't feature on the chart; however, the way I see it is that if the sensitivity of Luminol is at 1:5,000,000 and TMB is at 1:1,000,000, it's only a fivefold difference, which doesn't seem to be that much. The chart shows a dilution of 1:800,000 recorded at 361 RFUs, so I would put TMB at 1:1,000,000, recording at around 289 RFUs. You have a signal of 100 RFUs not being detectable by the naked eye, so we are entering the territory of very low luminosity when the Luminol reactions at VDP7 were quite vibrant. Vibrant enough for Garofano to conclude that they must be blood.

If TMB at a sensitivity of 1:1,000,000 recorded 289 RFUs at it lowest detection rate and no visible signal is recordable at 100 RFUs then the Luminol traces at VDP7 would have to be invisible or near to invisible, IMO for TMB not to detect it.

I'm no scientist, and I've muddled my way through this, so I'm looking for a better-informed consensus in layman's terms.

Hoots

I wonder if anyone has ever tried diluting a blood sample sufficiently to where TMB couldn't detect it, and then apply Luminol and photograph the results using the exact same settings on the camera as the SP used in the cottage. My expectation would be that the results would be significantly fainter than what we saw in the cottage, but that's just speculation.

 

[TomG]

I wonder if anyone has ever tried diluting a blood sample sufficiently to where TMB couldn't detect it, and then apply Luminol and photograph the results using the exact same settings on the camera as the SP used in the cottage. My expectation would be that the results would be significantly fainter than what we saw in the cottage, but that's just speculation.

Yes, that's a good point. You can see that the RFUs, according to all reagents, are diminishing to a very low luminosity at a 1:800,000 dilution. If TMB is still sensitive at a rate of 1:1,000,000, it means that it can still theoretically detect blood. Anywhere below the 1:1,000,000 dilution might mean no visual reaction since 100 RFUs are not even visually detectable.

Hoots

 

[Welshman]

Yes, that's a good point. You can see that the RFUs, according to all reagents, are diminishing to a very low luminosity at a 1:800,000 dilution. If TMB is still sensitive at a rate of 1:1,000,000, it means that it can still theoretically detect blood. Anywhere below the 1:1,000,000 dilution might mean no visual reaction since 100 RFUs are not even visually detectable.

Hoots

If negative TMB didn't rule out the footprints being made in blood, why did Stefanoni lie about the negative TMB results? Why lie to cover things up if the prosecution had a slam dunk case? In addition why did the prosecution never argue that negative TMB could still mean the presence of blood? There is something that PIP have overlooked which could be an effective pro innocence argument. If Amanda and Raffaele had left damming evidence such as bloody footprints, why didn't show they show any concern over this evidence?

 

[Numbers]

So, in other words, 6 of 9 Luminol revealed samples in the cottage, which tested positive for DNA, can't be blood based on this analysis.

How sure are you of your conclusion that in 2007, PCR positive TMB negative samples couldn't be blood? Are you aware of any forensic experts who have documented this?

Between this, the brightness of the Luminol reaction, and that Meredith's DNA profile was not found in 28 of the 31 Luminol revealed samples all suggest these samples were not made from Meredith's blood. But I'm not sure we have certainty about this relationship (TMB/PCR) nor am I sure we have all we need to properly evaluated the brightness of the Luminol reaction and what that means. I often accuse pro-guilt of seeing what they want to see, and I would prefer to not make the same mistake myself.

How many white blood cells (which contain DNA but no hemoglobin) are in a fixed volume of blood (say 1 ml) compared to the number of red blood cells (which don't contain DNA, but do have hemoglobin detectable by TMB or Luminol)? In other words, in a fixed volume of blood, what's the ratio of white blood cells to red blood cells?

Here's a generally accepted ratio from the Merck Manual*:

White blood cells (also called leukocytes) are fewer in number than red blood cells, with a ratio of about 1 white blood cell to every 600 to 700 red blood cells.

* https://www.merckmanuals.com/home/blood-disorders/biology-of-blood/components-of-blood

The point is that if the PCR is positive for human DNA and the presumptive blood test is negative, it is unreasonable to believe that the DNA came from blood. The DNA source would reasonably be some other tissue deposited by primary or secondary contact (for example, from skin cells), or it could be a contaminant.

I should follow up on how many white blood cells there are per a fixed volume of blood. This information is from the Cleveland Clinic**:

It is normal for you to produce nearly 100 billion white blood cells each day. After completing a blood draw, a test counts your white blood cells, which equals number of cells per microliter of blood. The normal white blood cell count ranges between 4,000 and 11,000 cells per microliter.

*https://my.clevelandclinic.org/health/body/21871-white-blood-cells

1 microliter = 0.001 ml; there are about 5000 microliters in 1 teaspoon (US).***

*** https://www.metric-conversions.org/volume/us-teaspoons-to-microliters.htm

Thus, there should be many thousands of white blood cells and their DNA in a small drop of blood.

 

[Numbers]

How many white blood cells (which contain DNA but no hemoglobin) are in a fixed volume of blood (say 1 ml) compared to the number of red blood cells (which don't contain DNA, but do have hemoglobin detectable by TMB or Luminol)? In other words, in a fixed volume of blood, what's the ratio of white blood cells to red blood cells?

Here's a generally accepted ratio from the Merck Manual*:



* https://www.merckmanuals.com/home/blood-disorders/biology-of-blood/components-of-blood

The point is that if the PCR is positive for human DNA and the presumptive blood test is negative, it is unreasonable to believe that the DNA came from blood. The DNA source would reasonably be some other tissue deposited by primary or secondary contact (for example, from skin cells), or it could be a contaminant.

I should follow up on how many white blood cells there are per a fixed volume of blood. This information is from the Cleveland Clinic**:



*https://my.clevelandclinic.org/health/body/21871-white-blood-cells

1 microliter = 0.001 ml; there are about 5000 microliters in 1 teaspoon (US).***

*** https://www.metric-conversions.org/volume/us-teaspoons-to-microliters.htm

Thus, there should be many thousands of white blood cells and their DNA in a small drop of blood.

Next, how much DNA does blood contain per unit volume:

When DNA is extracted from a whole blood sample, the vast majority of the recovered DNA comes from the white blood cells. The quantity can vary depending on the individual's white blood cell count (which can fluctuate due to health status, infection, or inflammation), but a typical yield for genomic DNA from a healthy adult is:

• Approximately 6 to 12 micrograms (µg) of DNA per milliliter (mL) of whole blood.

This amount refers to the high molecular weight genomic DNA extracted from the nuclei of white blood cells, which is commonly used for genetic testing, forensic analysis, and research.

Source: https://wellpath.life/how-much-dna-in-blood#how-much-dna-is-in-blood

The above source indicates (by calculation) that blood would typically contain 6 to 12 nanograms of DNA per microliter.

According to a 2024 source, the optimal minimum amount of DNA recommended by manufacturers for reliable PCR DNA profile analysis using 1995-era technology was 2 nanograms (ng); this minimum amount has decreased over time:

The optimal amounts of DNA recommended by manufacturers to generate probative STR profiles have substantially decreased, from 2 ng in 1995 ... to as little as 0.4 ng in current [2024] STR kits ....

Source:

PCR in Forensic Science: A Critical Review​

https://www.mdpi.com/2073-4425/15/4/438