Summary of Review of Labwork
Sorry for monster post, but I wanted to share:
The Lab Procedure
1. The lab utilized a highly unorthodox 50-cycle ICN (Increased Cycle Number) procedure in amplifying the samples via Real Time PCR (RT-qPCR). See
https://www.cps.gov.uk/legal/s_to_u/scientific_evidence/high_sensitivity_dna_analysis/ (stating that Crown Prosecution Service uses 28-cycle standard, and applies 34-cycle standard only secondarily and under restriction). The lab amplified all of its samples at 50 cycles, nothwithstanding the fact that the lab’s assay kit was validated for 28 cycles of amplification, the accepted standard for ordinary forensic samples is no more than 30 cycles, and the accepted standard for Low Copy Number (LCN) samples is 34 cycles.
2. In ordinary practice, a lab would subject to electrophoresis those samples that were positive for amplification product after completing 28 or 34 cycles (at which point the amplification reaction was halted). Stefanoni, however, amplified all of her samples at 50 cycles and used the CT (Cycle Threshold) value generated in the amplification process to determine which amplifications should be subjected to electrophoresis. The net effect of this unorthodox procedure was that samples were subjected to an additional 15 to 25+ cycles of amplification after exhibiting an acceptable CT value. In addition, some samples achieved an acceptable CT value well in excess of 28 or even 34 cycles. This procedure increased the likelihood of introducing otherwise undetectable, extraneous DNA contamination.
http://www.nfstc.org/pdi/Subject09/pdi_s09_m01_03_a.htm (noting likelihood of “undetectable, extraneous DNA contamination”); see also
http://www.denverda.org/DNA_Documents/LCN DNA Article Gill.pdf (discouraging amplification in excess of 34 cycles, because it “encourages artifact production”).
3. Generally, the lab subjected to electrophoresis those samples that amplified with CT values below 35. However, the fact that a sample amplified with a CT value below 35, and was therefore subjected to electrophoresis and generated an electropherogram, does not mean that the electropherogram was disclosed to the defendants. Indeed, a large number of electropherograms have been suppressed, and we can hypothesize that they reflect the expected outcome—i.e., contamination—of the unorthodox 50-cycle process employed by the lab.
4. Analyzing the lab results through November 23, 2007, for which amplification records have been produced (this excludes the samples that presumably were amplified in between November 6 and November 22, because the applicable PCR records have been suppressed) we can make the following observations concerning the creation and suppression of electropherograms:
a. CT value below 32: For samples that amplified with a CT value below 32, the prosecution suppressed only the electropherograms that were deemed damaging to the prosecution’s case. Specifically, known suppressed electropherograms with CT values below 32 include No. 601 (a 26-CT amplification of blood located in the downstairs apartment, belying the identification of downstairs samples as “cat’s blood”) and No. 939 and 944 (amplifications from Kercher’s bra, which would likely suggest contamination and therefore would support that the bra clasp was contaminated).
b. CT value of 33 to 35: For samples that amplified with a CT value of 33 to 35, the prosecution produced only those electropherograms that were helpful to the prosecution’s case (36b is perhaps included in this category); it suppressed all other electropherograms, including but not limited to Nos. 600, 602-04, 626, 628, 948 and 952, all of which would probably show contamination in the form of mixed/uninterpretable results, and therefore undermine the prosecution’s case.
c. CT in excess of 35: The lab did not process numerous positive amplification products in excess of 35 CT, because these samples likely would have generated mixed/uninterpretable electropherograms and therefore would have undercut the prosecution’s case by suggesting the presence of human DNA, possibly due to contamination in the field and/or lab.
5. The lab ran both negative and positive controls in at least part of its processes, however, almost all of the relevant records have been suppressed.
6. Each PCR run contained a set of negative controls (No Template Control (NTC)) and positive controls (based on processing of the “0.068 STD” lab standard). For the PCR runs for which records have been produced, the controls appear to have been successful; however, the PCR records for run nos. 545 through 548, which would contain the amplification of 36b, have been suppressed, and therefore there is no record of any control processed in conjunction with the amplification of 36b.
7. After processing via PCR, at least some of the control samples were subjected to electrophoresis. The positive controls can be located in gaps in the electropherogram production, meaning that the electropherograms were created but all have been suppressed, including but not limited to, electropherogram Nos. 622 and 631. The electropherogram records do not show an obvious gap for the negative controls.
8. All control records have been suppressed for the amplification and electrophoresis runs that produced 36b.
9. The ABI 3700 electrophoresis machine utilized by the lab suffers from known design defects known to cause contamination.
http://cstl.nist.gov/strbase/valida..._Ref 2340 FSS validation of 3100 and 3700.pdf
Batch One Results
10. Processing of the first batch of lab samples commenced on November 5 and was completed on or about November 6. Curiously, this batch is designated by a case number (L10747-01) that is different from the case number that was later used for the rest of the case (28669-01).
11. The results from the first batch show that the rapist and the intruder who left the feces in the toilet were one in the same person (i.e., Guede, identified at the time as Uomo No. 2), and that that person was not Sollecito or Lumumba. Thus, the prosecution had Guede’s DNA profile in hand as early as November 6 (probably prompting a concordant leak to the press that a “fourth” man might be involved). These exculpatory lab results, however, were not disclosed for the detention hearing on November 9. In fact, when the exculpatory electropherograms were subsequently produced, they were produced without date information, so that it remains obscured that the prosecution suppressed these exculpatory electropherograms in order to secure the pretrial detention order.
12. The first batch of lab results focused heavily on the downstairs crime scene, in particular, various blood spots that were collected inside and outside of the downstairs apartment. The lab successfully amplified human DNA in many of the samples from these spots. Further, some of the blood spot samples were subjected to electrophoresis, yielding electropherograms. Notwithstanding these facts, the downstairs/outside crime scene investigation was abandoned as soon as Knox/Sollecito were identified as “suspects,” and never revisited by the forensics squad. All samples of the downstairs/outside blood samples have been falsely described as “cat’s blood,” and all corresponding electropherograms (Nos. 600-604, 688-89) have been suppressed.
13. Because the lab was legally prohibited from conducting testing after suspects had been arrested, there was a 5-day delay in some testing for the period Wednesday, November 7 to Sunday, November 11. Notwithstanding this prohibition, it is evident that the lab in fact conducted ongoing testing on November 6, at a time after Knox, Sollecito and Lumumba had already been arrested.
Batch Two Results
14. The lab resumed testing on November 12.
15. It does not appear to be true that the RT-qPCR machine was out of service as of November 12, as was allegedly claimed; rather, we can identify that RT-qPCR run nos. 545 through 548 correspond to this period between November 6 (PCR Run No. 544) and November 23 (PCR Run No. 549). However, all records of RT-qPCR run nos. 545 through 548 have been suppressed, and therefore, there are no records of the amplification of the samples that were processed during the period November 12 through November 23.
16. Electrophoresis plate no. 365bis was processed on November 13, and contained sample 36b (kitchen knife blade). This plate is remarkable because (i) it is labeled with the suffix “-bis,” suggesting that for this plate, Stefanoni deviated from her ordinary procedures by using of bis-tris propane or similar agent, perhaps intended to “clean up” noisy test results, (ii) three out of the four disclosed electropherograms from this plate generated very low RFU readings, and iii) there are a large number of suppressed electropherograms from this plate, suggesting a serious problem with mixed and uninterpretable profiles.
17. The following egrams have been suppressed from the plate no. 365bis run: 758-60, 762 to 769.
18. The RFU scale for egram No. 36b suggests that it was generated from an extremely-low copy sample; perhaps so low that the profiled DNA could not have gone through a 50-cycle amplification process and therefore must have been introduced after amplification. Since the lab at some point in the process employed a concentration step, and since it appears that the samples in plate no. 365bis underwent some kind of “clean up” procedure, there may have been opportunity for such contamination after amplification but prior to electrophoresis.
19. Immediately after completing electrophoresis plate no. 365bis and isolating Kercher’s profile in sample 36b, Stefanoni created a second electropherogram of the sample, apparently as a single sample in the succeeding plate no. 366. This suggests that Stefanoni, herself, believed 36b to be the result of contamination, and possibly, that she had no control records to validate the original finding. It is unclear whether this subsequent 36b electropherogram was generated from the same or second amplification well for 36b.
Sample Storage and Recordkeeping
20. Many of the exhibits collected in the early days of the case were catalogued, packaged and stored together for a lengthy period, allowing for a significant opportunity for contamination if not packed correctly, e.g., packed in a calendar box.
21. Extensive laboratory documentation has been withheld. These records include: (i) evidence collection/receipt reports, (ii) logs/indices of lab traces, (iii) all authentic/contemporaneous extraction records, (iv) complete and authentic Qubit Flourometer records, (v) Real Time PCR records for run nos. 545 to 548, which would include the amplification records for 36b, (vi) records of concentration/purification processes, (vii) certain electropherograms (e.g., nos. 600-604, 617, 622, 626, 628, 631, 685-86, 688-89, 693-94, 758-60, 762-69, 939, 944, 948, and 952), and of course (viii) EDFs.
