I have a sense of deja vu, but hey ho. Let us try and lay this out as clearly as we can
The knife was collected from Sollecito's house. It was collected by a team which was not the same as the one which was at the cottage. The knife was selected from a drawer because it was noticeably clean and shiny. It was bagged and placed in a shoe box and sent to rome, where it was tested. Stefanoni also noticed that it appeared to have been very vigorously cleaned and so she took the decision to investigate it using non standard methods. She did not omit controls: she did not omit standard protocols. The test was witnessed by independent experts. The tests showed dna and that dna matched Meredith Kercher's dna.
The defence have two routes through which they can challenge that.
1. The test is not reliable and the dna is not in fact Kercher's: it is rather an artefact of the flawed protocol adopted
2. The test is reliable and it is indeed Kercher's dna: but it is a result of contamination in the lab
Knox's defence did indeed try to make both points: but the fact is that they have largely abandoned the first explanation. There is a reason for this. Greggy is a scientist working in the field and he had this to say at PMF
Greggy said:
Thanks ever so much for posting the graphs comparing the DNA results from the knife to the victim's wounds. Gosh, I love looking at actual data!!!
Every new protocol using a PCR amplification instrument has to be validated again and again before being accepted, but looking at those two superimposed graphs you posted and roughly figuring in standard deviations and area under the peaks by eye to assess significance, it is my hypothesis that the DNA on the knife and from MK's wounds are the same........ The pattern and its amplification doesn't perfectly match, but I never thought the marker data would be that good considering all the equivocations asserted in the press. I hope the Italian scientists publish their results so my lab guys can employ their methodology to achieve higher sensitivity. I only wish they hadn't farmed out the evidence to Applied Biosystems for the results (they're our biggest competitor).
in response to this Brian S wrote:
Brian S said:
In the early days of DNA testing because it was so little understood, the courts in various countries set what amounted to "arbitrary levels" that the peaks should exceed in order to be considered valid by the courts.
But science doesn't work according to man made rules, it just ain't that simple.
More and more the "arbitrary levels" set by the courts are being challenged. A very noisy sample with peaks over 60 which the court may consider valid is in many cases worse than a less noisy sample with lower peaks. The argument is on and cases are being challenged, even in the US, that the validity of DNA evidence should not judged on a subjective level set to avoid judges and juries having to make an effort to understand the science. The argument will be had. The suggestion is that wrong people have been convicted on noisy samples which exceed the court imposed levels whilst guilty people have gotten away with it because their peaks, clear amongst the low volume noise don't obtain the level required by the court.
and Greggy replies:
Greggy said:
The legal system didn't realize how quickly DNA data would be improved and set standards on this new methodology based on the results scientists originally were able to achieve. It is isn't the sheer magnitude of peaks, it is the signal-to-noise (S/N) ratio that is important - a direct comparison of the specific peaks to the background peaks. Generally in science, for a result to be considered significant you need a S/N ratio of at least 4:1, which is very roughly P<0.05 (i.e., the probability of the results being the same are 5%, or conversely, you are 95% confident there is a difference in the results). In brief, to be considered significant, a specific peak should be 4-times higher or have substantially more area than the background non-specific peaks. Before seeing the knife DNA results, based on press reports, I thought there may be spurious peaks due to artifacts from over-amplification or maybe even technician contamination of the evidence. The DNA results posted by Petafly evinces that is not the case. Although the novel methodology they used may be unvalidated, the results they obtained are impressive, ..... I applaud their ingenuity.
It seems clear that this has been understood by the defence and they, like everybody else who is directly involved in the case, accepts that this is Kercher's dna. The generalisations from "experts" who did not see all the evidence, and who have vested interests of various kinds, are not persuasive in this context. Nor are old introductory text books. The fact is that there is no real doubt. The signal to noise ratio is fine: the absolute height of those peaks is irrelevant. Harping on about those matters is at best a misunderstanding of their meaning.
The defence has now largely relied on the second possible line of attack: contamination. They have not shown any mechanism for that tp happen: nothing specific to this case at all. It is true that contamination can happen: but any dna evidence can be attacked in that way and so there has to be some evidence that it actually did happen before that is a serious issue. We have seen no such evidence
I am aware that those who are wedded to their position will not be swayed by this: they demand references and studies and will not accept the explanation from a working scientist I give above. But his explanation is in line with the paper I cited earlier and which Halides1 persistently misunderstands. There is nothing set in stone about peak height: nothing which makes 150 rfu or 40 rfu definitive: that is clearly established in the paper (written by some of those who wrote the "open letter" he is so impressed by)
I will also repeat that one of those who wrote that letter herself used non-standard techniques in the course of her legal work and that work was subsequently validated and is now widely accepted. This is what happens in rapidly developing fields
As I understand the science (and I am not a scientist) there is nothing in what has been presented here which serves to undermine the fact that that is Kercher's dna: and nothing to suggest contamination actually happened.
