Tuesday, July 13, 2010

DNA transfer in Strangulation

Part XIX in the Knox/Sollecito case

Update 1, 22 August 2010

I communicated with an anonymous forensic nurse about their domestic violence program. If an alleged victim complained of an attempted strangulation, complained of a partner’s grabbing their arm, or showed bruising or redness, they have been swabbing that area for DNA, for two years. Positive results in at least one case helped to convict someone. This information confirms the reasonableness of swabbing Meredith’s body in appropriate places, although the bruise on the nape of Meredith’s neck may have been the result of her being thrown against the wall (as some have theorized). We do not know whether swabbing the bruised areas was done or not.
____________________________


PM Guiliano Mignini’s reconstruction of the murder of Meredith Kercher was shown to the jury in the form of an animated video. In his speculation Amanda Knox grabbed Meredith Kercher by the throat and slammed her against a wall. The video superimposed actual shots of Meredith’s bruises with Amanda’s animated hand to imply that Amanda’s action produced the bruises. Later, Rudy Guede and Raffaele Sollecito held Meredith’s arms back, and Amanda stabs her (Barbie Nadeau, Angel Face, p. 160). In forensics professor Carlo Torre’s reconstruction of the crime, the single assailant grabbed Meredith by the throat and stabbed her. The area under Meredith’s chin and her nape were bruised; the former bruising was the result of the knife and the latter when the assailant put Meredith down.

Although Mignini’s theory of the crime suggests a much longer amount of contact between the assailants and victim than does Torre’s, both imply contact. None of Amanda Knox’s DNA was found in Meredith’s bedroom or on her body. Raffaele Sollecito’s DNA was found on a bra clasp (although the defense contests this piece of evidence) but not the bra itself. Rudy Guede’s DNA was found in several places, including on the sleeve of Meredith’s sweatshirt. Let us examine some forensic DNA studies to see if they shed any light on this tragedy.

The study by Wiegand and Kleiber gives a case study in which DNA evidence was collected 48 hours after a strangulation. They wrote, “Strangulation marks were clearly visible on the neck of the victim. Epithelial cells could be removed from the neck of the victim using separate cotton swabs for the left and the right side of the neck. Only the swab from the right side could be typed and included the pattern of the suspect (Fig. 3), a result which corresponded to the autopsy findings (the right neck side showed a higher intensity of bleeding in the muscles than the left side indicating a more intensive pressure against the right side). Altogether clear results could be obtained using four STRs (TH01, VWA, FGA, CD4) demonstrating the high utility and sensitivity of the method described.” These authors also conducted simulated strangulations, and they reported a success rate of better than 70%.

The simulated strangulation study in 2002 by Rutty used both SGMplus and LCN amplification. The simulated strangulation experiment was done with periods between the force and the sampling were 1, 5, 10, 15, 30, and 60 min, 2, 3, 4, 5, 6, 7, 8, 24, and 48 h and 3, 4, 5 and 10 days. When the author used SGMplus, he observed a full profile of the offender 7 out of 29 times, and always in the presence of the victim’s profile. When the author used LCN all 17 experiments yielded offender profiles, with the majority being partial profiles. Dr. Rutty wrote, “Of the test neck swabs, 19 yielded positive amplification results using SGMplus, 12 showed a victim-only profile and 7 a victim and offender profile with a full offender profile detectable up to 6 h after contact. When LCN was used (17 tests) all showed the offender to be present for all time periods i.e. up to 10 days. In the majority of cases it was a partial offender profile with the majority of the amplification result being a full victim profile.”

Dr. Rutty stated that “When considering the apparent time periods of DNA survival, passive transfer of the offender’s DNA onto the victim’s neck could also explain the presence of offender DNA several days after contact.” This study did not break down the results by time periods between simulated strangulation and DNA collection; therefore, it is difficult to draw firm conclusions based solely on the data presented.

A 2008 study by Graham and Rutty reexamined the question of DNA transfer to and from strangulation victims with an emphasis on innocent DNA transfers that might deposit DNA on the neck of a victim. In 24% of samples collected showed nonself DNA on the simulated victim from third party sources. The authors believe that such DNA transfers might confuse an investigation.

Finally, it might be helpful to return to the subject of primary and secondary DNA deposition. The study by Lowe and coworkers in 2002 looked at DNA transfer from a good DNA shedder to a poor DNA shedder to an object. By definition transfer was secondary from the good shedder, and transfer was primary from the poor shedder. When mixtures were observed, secondary transfer from the good shedder provided the major component, not primary transfer from the poor shedder. This study illustrates the dictum that one generally cannot infer the mechanism of how the DNA was deposited from the DNA itself. Along with the study by Graham and Rutty, this work suggests caution in the interpretation of nonself DNA on the body of the victim of strangulation. A similar caution should also be applied in the interpretation of DNA on the bra clasp.

In addition to the case study discussed by Wiegand and Kleiber, there are a small number of news reports of strangulations that mention DNA. One is from greater Detroit, and another is from Chicago. However, these articles do not specify the place on the body that was tested for DNA. In addition to the possibility that DNA was collected from the neck, it is possible that the victim’s fingernails contained the perpetrator’s skin cells or vice versa.

There are no reports that ILE found anyone’s DNA on Meredith’s bare wrists or her neck, and it is unclear whether or not the forensic police swabbed for DNA in these areas. It is also unclear whether some areas on the neck would have been free enough of Meredith’s blood to allow swabbing for the assailant’s DNA. Nevertheless it is difficult to see why the forensic police should not have swabbed the nape of Meredith’s neck or her wrists. If they did and found nothing, it would be strongly exculpatory, although it might fall short of proof of innocence due to uncertainties over collecting enough DNA or collecting it quickly enough after the murder. If the forensic police failed to swab these areas, it would suggest that they did not do as thorough a job as one would wish.

Bibliography
P. Wiegand and M. Kleiber, “DNA typing of epithelial cells after strangulation,” International Journal of Legal Medicine (1997) 110 :181–183. abstract

G. N. Rutty, “An investigation into the transference and survivability
of human DNA following simulated manual strangulation
with consideration of the problem of third party contamination,” International Journal of Legal Medicine (2002) 116 :170–173.
abstract

A. Lowe, C. Murray, J. Whitaker, G. Tully, P. Gill, “The propensity of individuals to deposit DNA and secondary transfer of low level DNA from individuals to inert surfaces.” Forensic Science International (2002) 129(1):25-34. abstract

E. A. M. Graham and G.N. Rutty, “Investigation into ‘normal’ background DNA on adult necks: implications for DNA profiling of manual strangulation victims.” Journal of Forensic Science, (2008) 53(5):1074-82. abstract

46 comments:

Rose Montague said...

Wow! All seventeen yielded a profile using the LCN DNA method. I guess it would be a good bet that if Meredith's neck and arms were tested using this method, it might show who was holding her in those places.

Of course, Stefanoni's new and unimproved LCN DNA testing procedure was limited to just one piece of evidence. Interesting.

Anonymous said...

Thank you, Chris, for your diligence in bringing to light all the many ways this case falls short. It's stupefying. What possible reason can there be for not having swabbed the bruises on this poor child in the search for the murderer? Yet another unturned stone.

Observer

Anonymous said...

http://www.injusticeinperugia.org/Appeal.html

My hot links key is missing, but cut and paste this link for more problems with the prosecution's case to be raised on appeal, including the argument that the DNA on the bra clasp was not Raffaele Sollecito's after all...this is definitely a question for you, Chris.

Observer

Randy said...

"If they did and found nothing, it would be strongly exculpatory, although it might fall short of proof of innocence due to uncertainties over collecting enough DNA or collecting it quickly enough after the murder. If the forensic police failed to swab these areas, it would suggest that they did not do as thorough a job as one would wish."

I can't say I wouldn't be surprised the forensic police did not swab Meredith's neck and arms.

odeed said...

halides,

Both news articles you linked do not state that the DNA found was from neck wounds.

You left out that the Wieglund 1997 study that PCR cyles were increased to 30-31, beyond the standard 28 cycles, which makes it LCN, something you have been critical of regarding the knife evidence, and Solecito's DNA found on the bra clasp was above LCN levels.

The Rutty/Graham paper 2008 concludes that more studies needed to be made before such evidence could be included in case work.

Also in the 2008 study, IIRC, there were only partial profiles matched on the neck swabs from unknown sources, and this is something you have also been critical of in this blog.

It seems to me that you and your "students of the case" have are less critical of any scientific evidence that you think may exonerate Knox and Sollecito,than the actual evidence which convicted them.

odeed.

Chris Halkides said...

Odeed,

I respectfully disagree with many of your comments. I was careful to note the lack of information in the news reports that I cited; it is not as if I were trying to pull the wool over anyone’s eyes. The 2008 article coauthored by Dr. Rutty was mainly concerned about the problems of third-party DNA innocently being mistaken for the perpetrator’s DNA. In my reading they were not saying that one would not detect the perpetrator as much as they were saying that one might detect someone else unrelated to the crime. Their work does not suggest that an investigator should not swab these areas, only that one should interpret the results cautiously. I am not certain why it is significant whether the unknown profiles are full or partial in the case of these studies. However, I believe that partial profiles present problems of interpretation not encountered in full profiles.

31 cycles of PCR is in between what is typically done in normal work (28 cycles) and what is often done in low copy number (LCN) work (34 cycles).

I am glad you mentioned the bra clasp. An anonymous commenter
at a JREF forum said this, “Copious amounts of DNA that resulted in a full profile that could only come about via direct and vigorous contact 'is' evidence.” There are a couple of problems with this assertion. One is that the amount of Raffaele Sollecito’s DNA was not especially large. Raffaele Sollecito’s appeal goes through the calculations and shows that the amount of his DNA is less than 200 picograms, which is one definition of LCN. Two is that trying to deduce whether any DNA sample arose from primary or secondary transfer on the basis of intensity of peaks or completeness of profile is contrary to the prevailing wisdom of the field of forensic DNA. See the article coauthored by Lowe for an example.

However, let us ignore those problems for a moment. It is known from other studies that a one-minute handshake is enough to transfer DNA from one person to another. The force involved in strangulation or in restraint is probably much greater; therefore, there is no reason not to expect DNA transfer in these events. The time and force in strangulation or restraint are surely equal to (and probably much greater than) the time and force Mr. Sollecito is alleged to have applied to the bra clasp. If one accepts the argument of the anonymous commenter above, would he or she not expect DNA transfer to occur to Meredith’s neck or arms?

Anonymous said...

Chris,

Interesting articles, even in the abstract.

Would the blood from the wound on Meredith's neck affect the recovery of DNA from the murderer(s)?

The jacket Meredith was wearing was swabbed and tested with the result of Rudy's Y haplotype being on the sleeve (not sure where on the sleeve). I believe the only other piece of clothing where Rudy's DNA was recovered was on her bra (I am not sure of that).

Whether Rudy did the murder alone or had accomplices wouldn't one expect there to be large amounts of DNA left on Meredith's clothing and body due to the severity of the attack, removal of clothing, etc.?

Christiana

Chris Halkides said...

Christiana,

I make the assumption that blood would either wash away DNA from the assailant or that its removal from the neck would also remove DNA from the assailant, but I have no citations from the literature. Dr. Carlo Torre’s reconstruction of the attack was that it was fast and occurred with contact only between the assailant’s non-knife hand and Meredith’s neck. Your question about how much DNA to expect is an interesting one. It would be helpful to have a list of where the forensic scientists swabbed for DNA and did not find it. The other unknown (at least to me) is whether Meredith removed any of her own clothes. I have been proceeding under the assumption that she did not, but I have no evidence, one way or the other.

odeed said...

@halides1

In the Weigand 1997 study, the cycles were increased to 30-31 because of low DNA levels. This study also only matched 4 STR loci (compared with 10 (+1 XY) for SGM+), and the hands and neck were cleaned beforehand.

As for Graham 2008, when you quote 24%, the full quote for that figure I can find:

"between 1 and 9 allele(s) of an unknown source were observed on five of 30 neck areas and seven of 20 finger pads sampled, accounting for 24% of all samples collected."

Now I need to go back and find how the 24% was arrived at, but a quick look at the tables show that full profile of the "victims" partner DNA was found in 2 out of 30 tests on the neck (there was one additional high level but partial partner profile), unknown/offender DNA was low and were all partial profiles (1-4 alleles) which was found on 3 of 10 victim necks.

And just to get the time frame correct, the 2008 paper was published after the murder, so I have no idea how you would expect the forensics team to know the conclusions in deciding whether they would swab for DNA.

As for Solecito DNA found on the the bra clasp, I have not seen definite source for the amount (I have seen 200pg, 1.4 ng and 200 cells), but the allele levels found were above 150 RFU threshold (closer to 200 RFU), which is acceptable in scientific studies and law enforcement labs in many countries including the USA.

LCN technique was developed/proposed for DNA amounts less than 100pg (look for Peter Gill 2000 Forensic Science International IIRC).

As for the one minute handshake, are you using the Oorshot study or the Ladd study, from memory both these studies cleaned the hands before testing (so not a real world test). The Ladd paper also contradicted the Oorschot study in it's conclusion, and Ladd found the peak level of DNA by transfer was about 50-75 RFU (Solecito was above 150 RFU), and no full profiles were found.

"..would he or she not expect DNA transfer to occur to Meredith’s neck or arms?"

The answer is no, there have been no other cases I know of where such evidence has been used to convict, and the papers you cite do not suggest that any DNA above LCN levels would be found, and probably consist of partial profiles IMHO.

And are you seriously equating skin on skin contact with the rough handling of a metal clasp in asking that?

odeed.

Chris Halkides said...

Odeed,

As van Oorshot (Nature, 1997) points out, washing one’s hands will lower the amount of DNA transferred, so in a real world setting there is likely to be more DNA than in a study which employs recent handwashing.

The total amount of DNA on the bra clasp was 1.14 ng. Raffaele Sollecito’s DNA was between 1/6th and 1/10th of the total, which puts it between 100 and 200 pg. This is discussed in Mr. Sollecito’s appeal.

The question of which definition of LCN to use is a complex one. Dan Krane, “Some of the problems associated with LCN (Low Copy Number) testing,” defined LCN as being less than approximately 500 pg of DNA. The Caddy report defines LCN using 200 pg as the threshold. Bruce Budowle and coworkers do not use a single numerical cutoff to define the LCN region: “However, because of variation in estimates of DNA quantity within and between methods of quantitation, LCN typing is better defined as the analysis of any results below the stochastic threshold for normal interpretation.” My point, however, is not to debate whether or not the bra clasp falls above or below the stochastic threshold at this time, but instead to point out that the amount of Mr. Sollecito’s DNA might be less than some commenters assumed.

odeed said...

As I said Van Oorshot was contradicted by the paper by Ladd 1999, and any lawyer would bring this up in court.

From memory, Caddy and the 2009 Court of Appeal, agreed that a grey area of between 100pg-200pg exists because of stochastic effects, however they go on to say that even within this range, reliable DNA profiles can be produced. As far as I am aware Sollecito's DNA produced levels above the accepted 150 RFU stochastic threshold from standard 28 PCR cycle amplification, calling it LCN using some definition, does not make it unreliable evidence.

All the studies you cite include investigation of low level partial DNA profiles, which required extra amplification cycles a day or two after transfer, and the lab methods chosen were to increase the probability of offender/victim transfer, it does not point to incompetence or a failure by the Italian forensic team that they did not perform such tests, as you want your "students of the case" to believe, especially as this is a subject matter still being studied with more recent articles as you should know.

odeed.

RoseMontague said...

My understanding reading the report is that Stephanoni only used 28 cycles on the knife. That would indicate to me that she felt she had enough material to get a DNA reading without doing a test for LCN DNA, which she proceeded to do using a method that is not supported by research. One conclusion that could be made is that the material did not contain DNA. She took 20% of that material and tested it for blood which was negative. One possible conclusion is that the material was not the type of material that would normally contain DNA (not flesh or skin for example).

Reading the appeals it would seen that other than the questionable collection methods documented by video, the main problem is getting an accurate Time Of Death. It seems that no body temperature was taken at the murder scene and the body was not weighed prior to autopsy. Common sense would tell me that this should be common practice and normal procedure. It would also seem to me that reference samples for luminol footprints should have included the other flat-mates, when you are dealing with some of these blurry blobs, I think it would be appropriate to have those handy. Not to mention no DNA samples taken from the other flat-mates.

I am no expert in this area but looking at numerous videos of the forensic team, including the one of the bra clasp collection I posted on the previous topic, it appears that this team did not act in a very professional manner.

Anonymous said...

I am no expert in this area but looking at numerous videos of the forensic team, including the one of the bra clasp collection I posted on the previous topic, it appears that this team did not act in a very professional manner.

Rose,

I have looked at those same videos and cannot make the same conclusions as you. There is not enough information.

An expert opinion(s) (other than those of posters on message boards - myself included) on the videos, including what are proper collection methods and what the Perugian forensic team did wrong, would hold more weight, in my opinion, than videos that have many breaks in time.

Also, transcripts of Stefanoni's questioning in court by both the defense and prosecution as to her collection and testing methods would be helpful as to whether any impropriety was done during the collection and testing of evidence at the crime scene.

Christiana

Chris Halkides said...

Christiana,

I respectfully disagree with you about one aspect of Dr. Stefanoni’s testimony. I have previously documented instances where forensic professionals testified to something that was shown to be factually incorrect (see “Tarnish on the Gold Standards,” by W. Thompson for one example). In other words, there is only so much stock I am willing to put into their testimony. On the other hand, one of the videos of the clasp shows the workers passing the clasp around and putting it on the ground. I have previously given citations that indicate that disposable tools should be used and handling should be minimal. The dirty gloves were also captured on video.

However, someone should ask Dr. Stefanoni whether or not her lab kept a written record of contamination events, as some recommend. What ILE ought to do is to see that the log files and electronic data files are finally released. At least that way the defense’s expert witnesses could see whether or not the lab followed its own protocols, as noted by Professor Hampikian.

Chris Halkides said...

Odeed,

Are you claiming that van Oorshot is incorrect with respect to the effect washing one's hands has on DNA transfer?

Anonymous said...

I have previously documented instances where forensic professionals testified to something that was shown to be factually incorrect (see “Tarnish on the Gold Standards,” by W. Thompson for one example). In other words, there is only so much stock I am willing to put into their testimony.

Chris,

I do put much stock in testimony in a courtroom. A witness can be questioned vigorously by the defense or prosecution and from that testimony both sides can bring in experts which dispute or agree with the forensic witness.

Documentation of other cases is helpful but is not the testimony of forensic witnesses in this case. I am interested in their testimony and how the defense and prosecution dealt with their testimony.

I will look at the reference you cite (Gold Standards). I am always interested in reading and learning.

Christiana

Chris Halkides said...

Italian police receive awards for their handling of the Kercher murder.

RoseMontague said...

Perhaps they should also receive an Academy Award for acting, a parade with sirens and horns blaring and plenty of confetti, and the keys to the Kingdom. Each can have a "framed" replica of that picture of Amanda that was hung up even before her verdict in the first trial.

Sometimes police criticism is fair and sometimes it is not. I don't think this is the appropriate way to handle that criticism in this case, even if some of the criticism has not been fair.

Chris Halkides said...

Odeed,

The first paper about DNA transfer during strangulation came out in 1997, and the second appeared in 2002. The fact that another paper came out in 2008 is not really germane to the question of what the investigators should have known.

It is worth bearing in mind that a simulated study of strangulation cannot match the combination of intensity and duration of an actual strangulation. I previously quoted the following passage from the study by Wiegand, “Only the swab from the right side could be typed and included the pattern of the suspect (Fig. 3), a result which corresponded to the autopsy findings (the right neck side showed a higher intensity of bleeding in the muscles than the left side indicating a more intensive pressure against the right side).” The above passage implies that if one encounters bruising on the victim (especially if one has reason to believe it arose from person-to-person contact), one should swab for DNA. Whether one will always find it is another matter. That is why my criticism of ILE was worded in hypothetical terms.

I will have to deal with the rest of your comments at a later time. Please note that I expect all commenters to maintain a civil tone in their remarks.

Anonymous said...

Italian police receive awards for their handling of the Kercher murder.

I wonder if the carabinieri were awarded for their meticulous handling of the computer equipment/hard drives.

Randy said...

"Swabs taken from Irene’s neck showed the presence of amylase, which indicates the presence of saliva. A forensic analysis of one of the swabs showed DNA consistent with two contributors—appellant and Irene."

http://webcache.googleusercontent.com/search?q=cache:AUtbO3fFFrsJ:www.courtinfo.ca.gov/opinions/revnppub/A117503.DOC+dna+neck+swab+strangle&cd=7&hl=en&ct=clnk&gl=us

if they do swab, the police may find something they can use...if they do not swab- they have nothing.

Chris Halkides said...

Odeed,

Let’s examine handwashing and its effect on completeness of recovery. Figure 1 in the 2002 study by Lowe et al. shows that the percentage of a full profile of DNA rises over the course of minutes to hours after handwashing. Good shedders show the most differentiation from poor shedders at 15 minutes. Therefore if a study of DNA deposition incorporates a handwashing step a few minutes before the DNA is deposited, it will probably result in less complete profiles than if no handwashing were done. This observation, coupled with the fact that the studies cannot match the combination of time and pressure of an actual strangulation, contradict your claim that “the lab methods chosen were to increase the probability of offender/victim transfer.”

The 1997 paper by Wiegand and Kleiber only looked at the CD4, FGA, TH01, and VWA loci. The relatively long period of time, 48 hours, between the strangulation and the finding of the body, might have decreased recovery in this actual case study. The 2002 study by Rutty used either 28 cycles of amplification or 34 cycles. The 2008 study by Graham and Rutty used 28 cycles, except when assessing shedder status.

With respect to LCN, I have some reservations with respect to the standard LCN technique. However, my main issues with LCN in the Knox/Sollecito case is that the knife profile was generated with an inferior version of LCN. Dr. Tagliabracci’s argument to the effect that Raffaele’s profile on the clasp was below at least one threshold for LCN raises the question of why it was not run at least twice, as is typical for LCN forensic typing.

odeed said...

halides1,

Van Oorschot 1997 found that the DNA recovered from surfaces touched was relatively high (upto 150 ng) and profiles were recovered using standard STR, Ladd 1999 repeated the study with the similar techniques but discovered that the DNA was low (upto 15ng) and that the profiles recovered were close to background noise and therefore uninterpretable.

odeed said...

Lowe 2002 found that full DNA profiles were recoverable one hour after washing, as far as I am aware non of the studies were immediately carried out after washing.

Further more, washing hands decreases the chance of mixed DNA profiles, and mixed profiles, as you should be aware, is a defense lawyers wet dream, with regards to DNA evidence.

Take Weigand 1997 study, 4 STR loci would be a probability of 1:60000 of an individual match (in the Italian population), Graham/Rutty 2008 found 1-4 alleles of offender DNA which would be a significantly less chance of a match than Weigand.

Include the problems LCN and stochastic effects when interpreting mixed profiles, then you have very weak evidence, which is why I am not surprised that you have yet to find any legal cases which would confirm your theory.

odeed said...

I note the hand waving away of the fact that this area still be studied. You even include Lowe 2002 which even states in the abstract "...further investigations are underway in order to further add to understanding of the issues of DNA transfer and persistence.", I have no idea how you expect the police to see into the future.

@Randy, it is well known that DNA can be obtained from saliva (it's actually common to find saliva around the face and neck area), but it isn't something which Halkides was suggesting, maybe he will in part XX of his classes, along with explanation of how the assailants licked Kercher's wrists to restrain her.

Anonymous said...

Anyone,

Is it true that mixed profiles are difficult to reference especially if they are of a male/female mixture?

It seems I have read (not sure where) that the female DNA masks the male DNA and more sensitive testing needs to be done to extract the male DNA from the mixture (and even that testing is not always successful in extracting a male profile). I do not know how difficult it is to separate mixtures when both are female DNA.

I'm beginning to think DNA of a crime scene is much more complicated and fragile than what I previously thought. While it is advantageous to have much DNA from the crime scene to point to a suspect does there exist the possibility you will have very little, even from a particularly brutal crime?

Christiana

Chris Halkides said...

Christiana,

In situations where the alleged victim is female and the alleged perpetrator is male, forensic scientists sometimes use YSTR
testing. YSTR testing specifically examines the DNA from the Y chromosome, to circumvent having to look for a small amount of DNA from the alleged perpetrator in a sea of DNA from the alleged victim.

Forensic DNA typing has a stronger scientific foundation that most subdisciplines of forensic science. However, there are many complexities that I have only begun to appreciate. It is interesting that about half of all forensic DNA profiles are either a mixture or are a partial profile. Experts do not always come to identical conclusions when analyzing mixtures, as I mentioned in a recent post. The analysis of mixtures is the subject of a number of scholarly articles. A good place to start might be the resources (pdf files and powerpoints) at www.bioforensics.com.

My impression is that the amount of DNA to be expected varies with the specifics of crime, among other variables.

Chris

Chris Halkides said...

Odeed,

If you examine the electropherogram of the knife, you will see that within any given locus, the two alleles differ in peak height by as much as 3-fold, when ideally they should be equal. The inequality of peak height is due to stochastic effects from the low number of DNA templates. Moreover, the peak heights are as small as about 21 RFU. Therefore, the argument that the forensic police did not swab Meredith’s wrist or neck because they did not want to deal with weak samples and/or stochastic effects does not cut much ice with me.

I was intrigued by your comment, “and mixed profiles, as you should be aware, is a defense lawyers wet dream, with regards to DNA evidence.” The bra clasp has DNA from several unknown individuals, yet the jury apparently still accepted it as good evidence. Does this not go against your claim?

RoseMontague said...

The DNA sample in the sink was also a mixture it appears: From Amanda's appeal:


"4. Finally, we must bring an additional scientific data that contradicts the
Grounds.
From reading the electropherograms for tracks mixed Knox - since Kercher
analyzed here can not exclude the presence of an additional track
Biological attributable to third female subject.
Dr. Stefanoni said:
Ø could not exclude a third person because it profiles
very balanced (May 22, 2009 hearing transcript, p.. 222).
Ø In that case there may be a third person
always female but the same features in
this mixture(May 22, 2009 hearing transcript, p.. 229).
Ø And she mixed in genetic profiles that relate to the Knox preclude us
was a third person? R - This is what I was trying to
say. I can not just exclude (hearing transcripts
hearing May 22, 2009, p.. 222).
Ø These couplings are different so that I can include other
people than those already present. This bit is u n 'and the say ...
Here of course we talk about compatibility because there are
certainly alleles of the victim and Knox, precisely with
this combination can not be excluded that there are
more ... (transcripts holding a hearing May 22, 2009, p..
225).
The presence of a third person was also represented by undetectable
consultant to plaintiff, Dr. Torricelli.
Ø here is a track mixed with the presence surely
at least two profiles, there are small peaks that can assume
possibly other appearances (June 6, 2009 hearing transcript, p..
101).

The possibility of a third female person might have
lead to Laura Mezzetti and Filomena Romanelli, the other two tenants
dwelling.
The Scientific Police, however, was not available to the genetic profiles of
two other young people.
The Stefanoni confirmed at the hearing:
Ø D and also had the DNA of other girls who lived in
house could also detect the DNA of a third or a fourth
girl who lived all saw in the house and used all the
same sink? A - If you could see the genetic profiles could
compare, that's me in advance I can not say, I can not suppose
(Hearing Transcript May 22, 2009, p.. 226)."

Chris Halkides said...

Odeed,

You are new here, and it is understandable that you might assume that this discussion forum is similar to PMF. It is quite different. I ask everyone here to treat each other and their arguments with civility and cordiality. You don’t have to agree with me or anyone else here, but sarcasm indicates disrespect. I have often referred to myself as a student of the Duke lacrosse case; the same term holds for the Cameron Todd Willingham case and the Kercher murder case. In my mind, all of us here are students of this case; no one here is the teacher. Yet as one of the finest scientists I ever met said, “All of us are smarter than any of us.” All of us can learn from one another; no one has the right to use vulgar or belittling language toward anyone else. As long as you act in accordance with these principles, your comments will be welcome here. I don’t ask people to leave because they disagree with me, but I reserve the right to ask that they leave if they are disagreeable.

The prosecution’s theory predicts that DNA might be deposited in specific locations on Ms. Kercher’s body, and I have suggested that they should have tested their hypothesis. Your argument boils down to the notion that ILE is not at fault if they did not swab: If they had, they might have generated a partial profile, LCN DNA, or a mixed profile. This position assumes the result before doing the experiment. You ignore the fact that some of the profiles in one study were complete, and you give no reason why a partial profile should be considered an uninterpretable result. Moreover, you continue to ignore the reasons why one might expect more DNA to be transferred in a real-world strangulation than in a simulation. In addition, as Randy correctly pointed out, if you don’t do an experiment, you have no data at all. I would prefer to have some data than no data.

You have misunderstood my concerns with respect to the LCN knife profile, many of which are based on the Johnson/Hampikian open letter. One problem with the knife profile is the intensity of the peaks; most are below 50 RFU, and some are as low as about 20 RFU. This is the lowest peak threshold I have ever seen. What would happen if all of the DNA profiles from this case used such a low threshold? A second problem is that LCN is typically performed in dedicated laboratory spaces with features designed to minimize contamination. Stefanoni gave no indication of taking special precautions in her work on the knife. As I have said before, LCN is typically performed twice or even three times to account for alleles dropping in or out. All of these are problems specific to the knife, not general problems of LCN profiling.

Chris Halkides said...

This New York corrections site
discusses the initial investigation of strangulation, partly from the point of view of the 911 dispatcher in cases of possible domestic violence. Under the section “medical documentation” it lists “DNA typing of epithelial cells.” It is difficult for me to believe that this refers to obtaining a profile of the alleged victim; therefore, it is more likely to refer to a profile of the alleged assailant. However, greater detail would be helpful, and I will pass it along if I learn something more.

odeed said...

In regards to the five profile found on the bra clasp, from what I have read the profiles on consisted of the full DNA profiles of Kercher, and the full profile of Sollecito, and I don't think that the defense disputed that.

There were also 3 other profiles, and these remained unidentified, I believe that the defense expert testified that one may have been from Knox, which suggests they were partial profiles.

There are plenty of science and legal articles covering interpretation of mixed and partial profiles, and the probabilities involved. Unfortunatley I did not save them for future reference, but one case study I remember was of the Arizona DNA offender database, where 9 loci produced 144 matches in a 65000 person database.

odeed said...

In regards to the double DNA knife, I don't think I have made any claims to the accuracy/reliability of the profiles found, if anything I am raising the point that you have previously opposed the LCN methods used to test the knife, which does not seem to be any different to the Wiegand 1997 study did in amplifying the DNA profiles.

I did not ignore your assertion that in the real world strangulation would leave more DNA than in the studies, you have not proven it. In fact the Wiegand study you cite simulated strangulation on the upper arm so more force could be applied, studies using inanimate objects found low DNA, other studies into assault found little or no DNA http://www3.interscience.wiley.com/journal/120716418/abstract, and the studies methods were to maximise DNA transfer but real world cases have more variables to be concerned with, which is why I am not surprised to see that there is legal cases you can find.

odeed said...

Finally, regarding the posts here and elsewhere, I actually joined PMF after posting my first few comments here, and I am aware of others (from either side of the debate) style and substance, but it has not influenced the way I post.

In one of my two posts (to date) on PMF forums, I described your blog being authoritative, and yourself as taking a teacher role and the readers as "students" (paraphrase), I should include that I have found your posts to lack any objectivity and on occasion hyperbole when criticizing prosecution.

As for my sarcastic reply to Randy, I realised after posting that maybe I should of rephrased it better, but I cannot apologise for it when you make comments such as

"The prosecution’s theory predicts that DNA might be deposited in specific locations on Ms. Kercher’s body, and I have suggested that they should have tested their hypothesis."

No where has the prosecution made this claim, and this is the first time you have suggested that they did. In your article you correctly summarize the prosecution theory that Kercher was held, and this is because of the nature of the wounds and lack of DNA under her fingernails.

This is blatant shifting the goalposts, blaming the prosecution for your theory, and was probably in the back of my mind that you would this when I posted the sarcastic comment.

Chris Halkides said...

Odeed,

No, the goalposts have not moved at all, but you may not be putting the pieces together correctly. The prosecution’s theory is what they showed in their animation: Amanda grabbing Meredith by the neck and Rudy and Raffaele later holding her by the arms, as described in the first paragraph of the blog entry. This version of the events indicates that DNA might transfer from the three individuals onto Meredith, as I said in my comment at 6:24 PM on 18 July. There is no getting around the fact that sufficient DNA for forensics might be transferred, especially when one considers the prosecution’s theory of an escalating assault. If DNA can be recovered from certain parts of Meredith’s body, why not other parts?

Many of your comments assume that ILE chose not to swab the nape or wrist area because of their knowledge that the profiles might be weak or mixed. In other words, you are assuming that ILE did know of the strangulation studies available at that time, a point that you have previously contested. If they did know, then their pessimism with respect to strangulation and restraint contrasts sharply with their optimism about the knife, when Stefanoni went ahead with the experiment when her DNA quantification test read that it was too low. But a further problem is your implicit assumption that these areas of Meredith’s body were not tested. What if they were tested and came back negative?

With respect to the bra clasp you leave out the important fact that Sollecito’s profile was between 5- and 10-fold weaker than Meredith’s. You also ignore the fact that Sollecito’s appeal (approximately on pages 144-145) is questioning whether all of the alleles are his. See the portion of the appeal document beginning with the sentence, “Finally, with regard to the interpretation of DNA profiles extracted from track 165B, the sentence has fallen into a serious and irreparable mistake.”

With respect to the analysis of mixtures, you did not comment upon the quotes I provided in a different blog entry. DNA forensic expert Peter Gill said, “If you show 10 colleagues a mixture, you will probably end up with 10 different answers.” With reference to [the Puckett case] case Dan Krane made an important point, “There is a public perception that DNA profiles are black and white. The reality is that easily in half of all cases—namely, those where the samples are mixed or degraded—there is the potential for subjectivity.”

Specifically with respect to the bra clasp, the exact identities of the other three contributors is important but so is the fact that their DNA is even there. If we accept that their DNA was deposited innocently, how can we be sure that Sollecito’s was not also deposited innocently? The same question should be asked in regards to the mixed profiles found in the sink, as Rose Montague noted above.

To be continued.

Chris Halkides said...

Odeed,

With respect to strangulation or restraint, holding someone so tightly that they cannot escape being stabbed with a knife is not the same thing as striking a child, which is the subject of the study in the abstract you cite.

With respect to LCN and the kitchen knife, most of the problems originated with the fact that Stefanoni did not know that the sample from the knife had so few copies of DNA. Therefore, she concentrated the sample and dropped the peak threshold from at least 50 RFU to 20 or fewer RFU, to compensate for doing only 28 cycles of PCR, not the typical 34 cycles. It is bad science to change your criterion in the middle of an experiment, and Rrudin and Inman’s “Introduction to Forensic DNA Analysis” (p. 121) specifically cautions, “It is important to have some predetermined limit to distinguish what is signal and what is noise.” In addition, one would have to perform negative controls at the same threshold for the experiment to have any validity, and there is no evidence that Stefanoni did this in the material given to the defense.

LCN tests are generally performed twice, and only those alleles showing up in both are scored. There was not enough material for this critical retest. LCN testing is typically done is special facilities equipped with positive air pressure, lighting that destroys DNA and other safeguards against contamination. Such facilities do not appear to have used in this case.

For the reasons given above, what Stefanoni did was not standard LCN, and another problem is that her technique has never been judged in a peer-reviewed publication. If she did attempt to publish a paper on this technique, other scientists would be able to critique her work. Thus there is no contradiction between accepting standard LCN while rejecting or criticizing Stefanoni’s homemade version of it. Dr. Mark Waterbury does a good job of summarizing these problems.

Finally, referring to this ordinary kitchen knife the “double DNA knife” is misleading. DNA cannot be interrogated as to when it was deposited. The presence of Amanda’s DNA on the handle means nothing, because there is an innocent and logical explanation for its being there: she cooked with it.

You did not understand why I brought up the Perugia Murder File discussion board. They have a good comment policy, but the site administrators don’t enforce it evenhandedly. I would not tolerate coarse language or rudeness directed toward you and more than I would tolerate it from you. Your attempt to rationalize your earlier sarcastic comments rests on a misinterpretation of what I said in regards to strangulation, as I discussed above. Even in cases where you are (or at least may be) correct, you could learn a thing or two from Rose Montague. When she disagrees, she takes care to do it in a manner that does not demean the other person. Finally, sarcasm is even less justified when a commenter has been wrong. A better course of action would be to concede certain points and move on to others where one has made a better case.

Chris Halkides said...

Odeed,

Naturally I am glad to have someone with whom to discuss the DNA forensics, even if it is someone with whom I disagree. I welcome different opinions as long as all remains civil.

Anonymous said...

Specifically with respect to the bra clasp, the exact identities of the other three contributors is important but so is the fact that their DNA is even there. If we accept that their DNA was deposited innocently, how can we be sure that Sollecito’s was not also deposited innocently? The same question should be asked in regards to the mixed profiles found in the sink, as Rose Montague noted above.

Chris,

What is known of the other three profiles on the bra clasp? How partial are they? I believe one has been somewhat attributed to Amanda but even that is in dispute. If it is possible that the profiles are very partial would not the fact that Sollecito's DNA was a full profile be important?

Christiana

Chris Halkides said...

Christiana,

The electropherogram that I saw was hard to interpret because the peaks were small. So I do not know how partial or complete they were. I do not think that one can tell much about the how the DNA was deposited by whether the profile is partial or complete. The studies I cited would seem to bear this out. Raffaele's profile is stronger than the partial profiles, IIRC, but it is also weaker than Meredith's; therefore, it is difficult to see why the intensity difference is important in one case but unimportant in another.

RoseMontague said...

What is known of the other three profiles on the bra clasp? How partial are they? I believe one has been somewhat attributed to Amanda but even that is in dispute. If it is possible that the profiles are very partial would not the fact that Sollecito's DNA was a full profile be important?

Christiana


What if one of those partial profiles could be attributed to Stefanoni's DNA profile or the DNA profile of one of the Lab Techs? That would certainly be a significant developement. Of course, if you only look for a match with the four reference samples you have (Amanda, Meredith, Rudy, Raffaele) it limits you as to the result. It is either one of theirs or unknown.

This is why I think the argument I hear stating that if contamination is not proven then it does not exist is a complete pile of rubbish. My point is you can't find contamination if you don't look for it, so avoiding contamination reports is fairly easy if you overlook the obvious things that may show contamination.

Anonymous said...

What if one of those partial profiles could be attributed to Stefanoni's DNA profile or the DNA profile of one of the Lab Techs? That would certainly be a significant developement. Of course, if you only look for a match with the four reference samples you have (Amanda, Meredith, Rudy, Raffaele) it limits you as to the result. It is either one of theirs or unknown.

What is known about the partial profiles? Were they degraded profiles (perhaps from being washed with other clothes)? If the partial profiles were from one of the forensic lab technicians would they not be more than partial profiles? And do we know that they were not compared against the profiles of the lab technicians?

This is why I think the argument I hear stating that if contamination is not proven then it does not exist is a complete pile of rubbish. My point is you can't find contamination if you don't look for it, so avoiding contamination reports is fairly easy if you overlook the obvious things that may show contamination.

That is why it would have been important for a representative of the defense to have been at the lab during the testing to see that proper procedures were followed. So we rely on what was testified to in court. I have no reason to believe Stefanoni lied in her testimony concerning procedures that were followed.

Christiana

RoseMontague said...

Christiana,
I posted a quote regarding this in the next thread regarding the only two female reference samples she had (Amanda and Meredith). You can take that as may be, I have seen this elsewhere just unable to pin it down at present.

Sefanoni has testified and I think it is clear from her testimony that standards and protocols were not followed for testing LCN DNA. The procedures she followed were completely contrary to protocol and her testimony only confirms this.

The knife as the murder weapon only works if there is another knife or two involved which has not been proven. The prosecution theory that Amanda carried this around in her bag for protection has not been proven or even shown that it is somewhat likely. The idea that they would get rid of all the other evidence and yet put the knife back in the drawer to use the next time they cooked is a bit whacked.

As far as the bra clasp goes the video I posted is convincing to me that they "expected" big things from it. That is purely subjective but the way it was left for weeks, moved around somehow, and handled during this video showing the return for this evidence does not inspire me with much confidence in the veracity of the result obtained.

Chris Halkides said...

Christiana,

I think it is very difficult, perhaps impossible, to predict when a partial profile shows up and when a full profile shows up. The trouble with relying on testimony is that you are relying on someone's memory, which may be faulty. I would rather see laboratory notebooks and various types of files.

Anonymous said...

I think it is very difficult, perhaps impossible, to predict when a partial profile shows up and when a full profile shows up. The trouble with relying on testimony is that you are relying on someone's memory, which may be faulty. I would rather see laboratory notebooks and various types of files.

Chris, you and I are both in agreement here. I would definitely like to see laboratory notebooks and various types of files.

I would assume that when Stefanoni testified in court she had access to that information via her laptop or notebooks so hopefully she didn't have to rely on memory.

Christiana

RoseMontague said...

"The problem in discussion was indeed that according to Tagliabracci the DNA on the bra clasp was not enough to have a reliable test. And Stefanoni suggested Comodi to say that instead it was the perfect quantity, being 1.4 nanograms. An ideal quantity. But that revealed a problem: why nobody else knew that measure?"

Stefanoni has info the defense did not get

It makes me wonder what other, possibly exculpatory information she is still holding back.

Chris Halkides said...

To all,

I updated this post after discussing these issues with a forensic nurse. Her comments strengthen the case that swabbing for DNA is justified in cases of attempted strangulation.