A reconsideration of the DNA forensics in the Duke lacrosse case

Updated 5/18/2015

Introduction

In the spring of 2006 three players on the Duke lacrosse team were indicted for rape and kidnapping.  The accuser claimed to have been raped by three men at least one of whom ejaculated inside of her body.  Some DNA results emerged in the spring of 2006. Eventually additional evidence was uncovered, and the overwhelming perception that this evidence was exculpatory brought about a major turning point in public opinion prior to the declaration of innocence.  After evidence of his misconduct emerged, the district attorney withdrew from the case, which then fell to the office of the Attorney General of North Carolina to investigate.  After the investigation was complete, the Attorney General declared the three players to be innocent in the spring of 2007, and the criminal matter ended.

The Duke lacrosse case is generally perceived as a highly visible case of prosecutorial misconduct and a rush to judgment by the press and the general public.  However with the 2014 publication of William Cohan’s book The Price of Silence, this view has now been challenged, and Mr. Cohan’s view of the DNA evidence is at odds with many who have commented on the case.  Therefore, a reevaluation of the DNA evidence is warranted on the basis of a general understanding of this case, as well as other cases in which the DNA evidence plays a central role, such as the Knox/Sollecito case, a frequent subject of this blog.

The initial results from the rape kit:  Is the absence of evidence evidence of absence?

The initial results from the NC SBI laboratory did not find semen, blood, or saliva.  Subsequently Y-chromosomal (Y-STR) tests were performed at a private laboratory.  Investigators found no DNA from any Duke lacrosse player in the rape kit items but did find DNA from Mangum’s boyfriend.  DNA from plastic fingernails was also extracted.  Some of the results from the private lab were released in a brief report issued on May 10, 2006, but other results were withheld.  Apart from the plastic fingernails (discussed below), how significant was the lack of DNA?

One blogger posed questions along these lines to several highly qualified experts, including Dan Krane, Elizabeth Johnson, and William Thompson.  Dan Krane responded: “It is worth noting that DNA tests are amazingly sensitive (DNA profiles can be generated from as little material as that left behind in a fingerprint) and Y-STR tests have the potential of determining a male’s DNA profile even when a female’s DNA is present in hundreds or thousands of times greater quantities. Scientists are always wary of asserting that the absence of evidence is not proof of absence but it certainly is reasonable to expect to find a rapist’s DNA associated with a victim when the victim presents herself to investigators within hours of an attack and when she has not bathed, the rapist did not use a condom and ejaculation occurred.”
  These experts gave thoughtful, nuanced replies. The consensus view seemed to be that the degree to which the absence of evidence was evidence of absence lies in the particulars of each case.

What then are the particulars of the Duke lacrosse case?  A site at Reed College lists the following guidelines for preserving evidence of a possible sexual assault:

 “Do not shower or douche

Try not to urinate. Urinating may reduce the ability to detect “date rape” drugs

 If there was oral contact, do not smoke, eat, or brush teeth

 Do not change clothes. If you have already changed your clothes, place them in a paper bag (plastic may destroy evidence) If you haven’t changed, keep the original clothes on and bring an extra set to wear home from the hospital

Go to a hospital with the capability of providing a SAFE exam and request the exam. The cost of a SAFE examination is paid for from a state fund.”

The alleged victim in the Duke lacrosse case was taken into custody shortly after the incident and did not have the opportunity to shower or to change clothes.  Therefore, the chances of preserving evidence would have been high, if an actual assault had occurred.  William Anderson wrote, “Furthermore, no one — no one — who inspected the room (and police gave that bathroom a thorough inspection, taking swabs everywhere and testing for DNA) found any evidence of bleach products having been recently used. They found no DNA of Mangum, and none of [Collin] Finnerty and [Reade] Seligmann, and nothing else that demonstrated that either of those two young men had ever been in that bathroom.”  Parenthetically, Taylor and Johnson reported that no fingerprints belonging to Mr. Finnerty were found in the bathroom (Until Proven Innocent, p. 183).

Unfortunately, the question of whether or not the accuser said that the victims used condoms has been not entirely without controversy.  Stuart Taylor and KC Johnson wrote (Until Proven Innocent, p. 327), “In her report from the night of the alleged attack, [SANE nurse Tara] Levicy had written without qualification that Mangum had said repeatedly…the rapists had used no condoms and that she was sure that at least one had ejaculated—in her mouth.  But on January 10 [2007, nine months later], according to [Linwood] Wilson, Levicy stated that Mangum “said ‘no’ but wasn’t really sure” whether her attackers had used condoms.”  Even if one finds Ms. Levicy to be credible about Ms. Mangum’s lack of certainty (which I do not), DNA may be transferred by simple physical contact, as is borne out in studies of simulated strangulation.  In addition, perspiration contains DNA, and sebaceous fluid is a probable source of touch DNA, according to a study by Vecchiotti, Filippini and their coworkers.  Thus even when a condom is used during a sexual assault, there may be transfer of DNA.

Asked whether or not the use of a condom could have been detected, Elizabeth Johnson responded, “Testing for these substances is not typically done, despite what you see on CSI. There has been some research done re spermicides on condoms, but none of this is done as part of a typical test and validated methods for casework aren’t there yet.”  Of course, even if condom use were detected, it would not differentiate between consensual and nonconsensual sexual activity.

The lack of body fluids is absence of evidence in its own right.  It also weakens the DNA evidence, meaning that a given DNA profile found in the absence of a particular fluid might have arisen from any biological tissue or fluid.  Peter Gill describes framework or hierarchy of levels (Misleading DNA Evidence, p. 19) at which DNA evidence can be evaluated:

1.     The sub-source level refers to the strength of evidence of the DNA profile itself.

2.     The source level is an evaluation of the strength of the DNA profile if it can be associated with a particular body fluid, such as semen or blood

3.     The activity level associates the DNA with the crime itself, e.g. sexual assault.

4.     The highest level deals with the ultimate issue of guilt/innocence.

Professor Gill’s book gives examples of cases in which a forensic scientist attempted to claim a higher level for the evidence than was warranted, based upon principles of forensics and probability.  The DNA of the rape kit is sub-source DNA, in this categorization.

The lack of DNA from the accused coupled with the lack of body fluids and the lack of other evidence in the bathroom is strongly exculpatory.  Whether or not it constitutes “proof of innocence” cannot be determined without defining the term.  However, I would weight this lack of evidence more heavily that eyewitness testimony, even if the latter came from a credible witness, owing to the well-known problems with eyewitness testimony.  Ms. Mangum’s credibility is questionable on a number of grounds.

The plastic fingernails in the trashcan

Ms. Mangum’s painted fingernails were placed in the trashcan of the bathroom and the players pointed them out to the police investigators days later.  David Evans could not be excluded as a donor.  The summary of conclusions from the office of the Attorney General of North Carolina stated on p. 12, "First, statistically, the chance of randomly selecting an individual from the population that could be included in this sample would be approximately 1 in 1000." With respect to the autosomal (standard DNA) profile, David Evans (along with about 2% of the population) cannot be excluded as a donor.  Likewise in the Y-chromosomal (YSTR) testing, David Evans cannot be excluded as a donor.  Although the summary report did not explain the calculations, the value of 1 in 1000 probably derives from combining the data from the two types of DNA testing.

The author of a 2014 book on the case (The Price of Silence), William Cohan clearly thought that the DNA was highly incriminating, discussing it on pp. 277-278 and pp. 325-326, among other places.   In the final chapter, Denouement, he returned to the subject on p. 602:  “What remains unresolved is, if in fact it was David Evans’s DNA on Mangum’s red plastic fingernails, how did it get there?”

DNA evidence beneath real fingernails typically is probative, not necessarily conclusive, evidence against a defendant.  Studies have shown that mixtures of DNA under fingernails are seen about 5% of the time (Gill, pp. 43-46).  However, there is less to the DNA from the plastic fingernails than meets the eye.  Taylor and Johnson wrote (Until Proven Innocent, p. 221), “And it would later become clear…that the DNA that might (or might not) have come from Evans had not necessarily been lodged “under” a plastic fingernail.  Rather, the SBI had extracted and mixed together all of the DNA taken from all surfaces of the three used plastic fingernails found in the wastebasket in Evans’ bathroom.”  Moreover, a case in Britain involving a taxi driver, David Butler suggests some additional caveats.  His DNA was apparently found on the fingernails of a murdered woman Anne Marie Foy.  He was convicted partially on the basis of the DNA evidence but later released.  A plausible theory is that he handled money which later carried his DNA to the victim.  There are not many forensic details that have been made public, but a few things can be gleaned from the information that is available.  One is that Mr. Butler's skin was unusually flaky.  Two is Hannah Barnes' report that "The victim was also wearing a glitter nail polish, which proved particularly attractive to dirt - and DNA." 

The defense correctly noted that David Evans was not uniquely identified in either of the tests.  The defense’s position was that if David Evans were a donor of the DNA, then there existed a plausible route for innocent DNA deposition, namely secondary transfer from materials in trashcan, such as dental floss and tissue paper.  Defense attorney Joseph Cheshire, "Any expert and any person in the world will tell you that your [own] DNA is in your bathroom." (Cohan, p. 242).  Saliva is a good source of DNA, for example.  The office of the attorney general of North Carolina accepted the possibility of secondary transfer as a reasonable explanation.  In the summary document pertaining to this case, the authors wrote: “Third, to the extent that Evans’s DNA could not be excluded, the SBI experts confirmed that the DNA could easily have been transferred to the fingernails from other materials in the trash can.”

It is uncontroversial that the presence of DNA does not indicate the time or manner of its deposition.  This cuts both ways: one cannot be certain that the DNA on the plastic fingernails arrived by secondary transfer, but neither can one rule it out.  Secondary transfer in DNA forensics is well established, and tertiary transfer has been observed under some circumstances.  Neither the amount of DNA nor the completeness of the DNA profile is a reliable guide as to whether the DNA arrived by primary transfer or secondary transfer.  Joe Minor and Suzanna Ryan have written several well-researched and readable articles that cover the subjects of DNA found on touched objects and secondary transfer.  Review articles by Van Oorshot and colleagues and by Meakin and Jamieson in academic forensic journals may be consulted for more detailed information.

Shedding and transferring DNA is perhaps more common than the general public appreciates.  Leslie Prey wrote, "We all shed DNA, leaving traces of our identity practically everywhere we go. Forensic scientists use DNA left behind on cigarette butts, phones, handles, keyboards, cups, and numerous other objects, not to mention the genetic content found in drops of bodily fluid, like blood and semen (Van Oorschot & Jones, 1997)."  The issues of shedding and transferring DNA were perhaps not very important when DNA profiling was in its infancy.  However, the introduction of techniques based on the polymerase chain reaction (PCR) meant that much smaller amounts of DNA are needed now than in the first generation of DNA profiling.  This is because the DNA is increased in amount (amplified) many times over prior to the analysis.  The more recent development of low copy number DNA profiling means that still smaller amounts of DNA can be typed, less than a dozen cells.

Although secondary transfer from the waste in the trashcan is highly plausible, primary or secondary transfer via the players themselves is also possible.
  Joe Minor wrote, “On one occasion, I swabbed my own hand after handshakes at a social function to determine the presence of other epithelial cells.  The laboratory was able to obtain a mixture of my DNA as well as two other individuals.”  Suzanna Ryan wrote, “One study performed by Lowe, et al., was designed to highlight a ‘worst case’ scenario and involved two individuals. The first was determined to be a poor shedder and the second a good shedder. These two shook hands for one minute. The poor shedder had washed their hands immediately prior to the experiment whereas the good shedder had not. After shaking hands the poor shedder held a sterile plastic tube for 10 seconds. The tube was then swabbed and tested for the presence of DNA. This experiment was performed on two sets of good shedder/poor shedder pairs. Surprisingly, in one of the pairs, only the good shedder's DNA was obtained from the plastic tube, with no evidence of a mixture including the poor shedder!”

  Results from Vecchiotti’s laboratory published in 2014 are consistent with this result.  Suppose that David Evans shook hands with Ms. Mangum when she arrived. That or a number of other events, such as handling money, might also transfer DNA.

  As a team captain and resident of the house, Mr. Evans might have shaken hands with Ms. Mangum when she entered or touched money that she later handled.

As persuasive as these arguments are however, one even more powerful argument may also be offered in Mr. Evans’ defense using the tables that Professor Giannelli presented in the chapter “DNA Profiling” within the book Race to Injustice.  At least two other males contributed DNA to the sample (not lacrosse players nor other men who attended the party).  In the YSTR profile ten of sixteen loci (locations within the Y chromosome) show the presence of three contributors and two more loci show the presence of at least two contributors.  In the autosomal profile some loci (places within other chromosomes) clearly show that at least three individuals contributed DNA.  This is very problematic for the prosecution.  If one acknowledges that this DNA arrived in a way unrelated to a sexual assault, then how is it possible to exclude the possibility that Mr. Evans’s DNA also arrived innocently?

By itself, the presence of what may be Mr. Evans’ DNA on the fingernails seems inculpatory; in the context of the other profiles and given the strong possibility of secondary transfer by multiple mechanisms, it is close to meaningless as evidence.  Paul Giannelli summed up the DNA on the plastic fingernails: “There was some probative value--but not much.”  Followers of the Knox-Sollecito case will recognize that the presence of additional male DNA donors is basically the same situation that exists with Mr. Sollecito’s Y-chromosomal profile that is associated with the bra clasp.  The Gary Leiterman case is perhaps the starkest example of this conundrum for the prosecution; the additional profile belongs to someone who could not possibly have been present at the crime scene.

Two more issues deserve attention, one of which is the difference in how the SBI versus DNA security interpreted the mixture in the autosomal DNA profile.  Taylor and Johnson wrote (Until Proven Innocent, p, 221), “The SBI had reported to Nifong that there were ‘multiple contributors’ to the DNA mixture but that ‘no conclusion [could] be rendered’ as to whether any of this fingernail DNA was even a possible partial match with any of the players…Based on the same testing process that had led the state lab to find no basis for any conclusion at all, DNA Security offered a singularly weak conclusion: that 98% of the male population could be ‘excluded as a contributor to the mixture’ and that David Evans was in the other 2 percent.”  Section 21C of the motion to compel discovery also quotes the same portion of the SBI report, and footnote 70 emphasized that the SBI had said “that no conclusion ‘could be rendered as to the contribution of DNA profiles from the buccal swabs of the suspects’ in the minor profile BS 4522.” 

The defense was rightly concerned about this apparent discrepancy.  It is possible that DNA Security used a suspect-centered analysis, meaning that they analyzed the mixture with Evans’s reference profile in hand.  However in “Painting the target around the matching profile: the Texas sharpshooter fallacy in forensic DNA interpretation,” William Thompson observed, “In the absence of clear standards for ‘inclusion’ and ‘exclusion’, different experts evaluating the same evidence may reach different conclusions: one may conclude that a particular suspect is ‘included’, while another concludes that the same suspect is ‘excluded’.”  Indeed, Dror and Hampikian have shown that the same mixed DNA profile may be interpreted differently by different groups of analysts.  Unfortunately the interpretation of DNA mixtures is presently still a somewhat subjective process; therefore, the difference between the two labs may have an innocent explanation.  However, another take-home message from this case is that the defense should re-analyze DNA mixtures from items of evidence.

Finally, Dr. Meehan’s own DNA may have contaminated an item of evidence.  Such contamination events are not uncommon.  The ABA’s standard 4.1a states in part:

“(ix) reports of laboratory contamination and other laboratory problems affecting testing procedures or results relevant to the evaluation of the procedures and test results obtained in the case and corrective actions taken in response…”

If there were a contamination event, Dr. Meehan’s laboratory should have reported it.  However, laboratories do not always do so.  In the Adam Scott and Gary Leiterman cases, the laboratories simply ignored evidence of DNA contamination in the negative controls (negative controls are experiments which should have no DNA in them).  There are also instances in which negative controls were faked in some way, as noted by William Thompson in “Tarnish on the gold standard.”  The reasons to fake having performed the negative controls include covering up a contamination event and increasing laboratory throughput.  Given the existence of DNA contamination and of instances of samples being mixed up or mislabeled in some way, it is surprising and disappointing that anyone opposes complete disclosure of the raw data, contamination logs, standard operating procedures, and any other pertinent records.

The DNA evidence released in the fall

On 27 October 2006 Nifong released some 1800 pages of material to the defense, but still did not include a complete report.  Although he lacked training in this area, defense attorney Brad Bannon did an extensive study of the material and was able to glean the essential information that the Y- chromosomal DNA of four unidentified men (men who were not at the party or Mangum’s boyfriend) had been found in the items of evidence from the rape kit.  When this information become public knowledge, it was a major turning point in the case, one that helped to shift public opinion in favor of the three accused students.

These profiles are important for at least three reasons.  One is that the finding of DNA serves as evidence that the tests were working, in other words the results functioned as a positive control.  The technique of amplifying DNA using the polymerase chain reaction (which is central to modern profiling) is susceptible to inhibitors of DNA polymerase, the enzyme that copies the DNA.  The lack of a DNA profile might in principle, arise from the presence of inhibitors, as opposed to the absence of DNA.  Of course, it is likely that the lab ran positive controls, but this sort of information might not come to light unless a DNA expert had reviewed the case for the defense.  Two is that Ms. Mangum’s sexual activity provides an explanation for the edema noted in her examination.  Three is that Ms. Mangum had claimed no sexual activity for a week, and these results raised doubts about her credibility, which would have been a central issue if the case had come to trial.

Using the information that they had obtained a month and a half earlier, the defense teams jointly crafted a motion to compel discovery and filed it on 13 December 2006.  Besides the information gleaned above, the defense teams learned that DNA Security had tested more items than it had disclosed in its report:

There are significant gaps in the item sequence numbers…The worksheet shows that the DNA extraction was performed n July 12, 2006, on that item, as well as items labeled “16074C” and “16075C,” which are nowhere otherwise identified in the DNA Security materials.  But on this worksheet itself, under the “name” column, someone typed “Bobby” and “Owen.”  “Quantification Worksheet Q-181”⁷⁵ and “PCR Worksheet Number A-464”⁷⁶ reflect quantification & amplification work on those three extractions the following day, July 14, 2006.  “Analysis Worksheet Number G-592” reflects analysis of those extractions the following day, July 14, 2006.  But nowhere in the materials do the DNA profiles generated from those analyses appear.  The phenomenon repeats itself throughout the underlying materials from DNA security.⁷⁷

Students of the Knox/Sollecito case are familiar with this kind of withholding of evidence, except on a larger scale; there are many samples for which DNA quantification was performed but for which no electropherogram (loosely speaking, an electropherogram is a picture of the sizes of the DNA fragments) was produced.

DNA Security’s standard operating procedures with respect to its reports were shown to be in marked contrast with its conduct in this case.

52.  Included in the materials provided to the Defendants from DNA Security are the laboratory’s standard operating procedures, which include guidelines for drafting “Reports and Documentation” of analyses performed by the lab in each case:

Reports shall include:

·      Case identifier.

·      Description of evidence examined.

·      Description of methodology.

·      Name of each DNA locus analyzed.

·      Results for each DNA test.

·      Interpretative statement of conclusions…

That the results for each test should be part of the report is very clear.  Yet Dr. Meehan chose to disregard this standard for reasons that are open to speculation.  Perhaps he acted as he did because he wished to establish a business relationship with the office of the district attorney or because of the publicity that working on a high-profile case would bring his firm.  The defense motion also noted, “the troubling effect of its limited scope of reporting is that it allowed DNA security to produce a report in this case that avoided disclosure of exculpatory evidence, either in the form of potential contamination noted in its testing, or as it relates to the discovery of DNA from multiple male sources…”  It is difficult to overstate how different the actual results are from the results as portrayed in the incomplete report from DNA Security produced on 12 May 2006.

DNA laboratory reports and the discovery of DNA evidence

The American Bar Association’s model rules make several important points regarding DNA evidence, one of which also insists that the reports be complete.  Standard 3.3a states, “(a) A summary of all DNA testing and data interpretation should be recorded promptly in a report.”  Standard 4.1a states in part, “(xi) material or information within the prosecutor’s possession or control, including laboratory information or material, that would tend to negate the guilt of the defendant or reduce the punishment of the defendant.”

Robert Mosteller discussed Nifong’s transgressions: “Nifong had an ethical duty under Rule 3.8(d) to provide the exculpatory information in a timely fashion, which is not explained further by the rule, but lacks any suggestion that prolonged unjustified delay is authorized.”  Kenneth Williams concurred, writing, “By withholding exculpatory evidence from the defense, Nifong violated the U.S. Constitution, the laws of North Carolina, and Rule 3.8d of the North Carolina Rules of Professional Conduct.  Specifically, the State Bar found that ‘Nifong did not make timely disclosure to the defense of all evidence or information known to him that tended to negate the guilt of the accused’ and that he ‘failed to make a reasonably diligent effort to comply with a legally proper discovery request.’”

Professor Paul Giannelli commented on the incomplete laboratory reports, “In any event, no attorney should have to search through the haystack for the exculpatory needle.  A laboratory report should be comprehensive and include a section specifying the limitations of the technique used in the analysis.  The report should also be comprehensible to laypersons.” (emphasis mine)  Professor Giannelli noted that Dr. Brian Meehan also failed to meet the standards for reports laid out by the American Society of Crime Directors/Laboratory Accreditation, even though his laboratory relied upon this organization for accreditation.

Conclusions

The circumstances surrounding the absence of evidence coupled with the presence of DNA from other men is persuasive evidence of absence; no one who attended the party sexually assaulted Ms. Mangum.  The DNA that may be from David Evans on the plastic fingernailsis very weak evidence against him; in isolation it would raise the possibility of his participation in a sexual assault.  However, there are plausible alternative explanations for how his DNA arrived; therefore, if this were the only piece of evidence in the case, it would not come anywhere close to the threshold of reasonable doubt.  In the context of the absence of Mr. Evans DNA in the rape kit and the presence of other men’s DNA on the plastic fingernails, it loses almost all significance.

Regrettably, the withholding of evidence is a common theme in miscarriages of justice; however, two factors were not in Mr. Nifong’s favor.  In addition to North Carolina’s open discovery law, the defense would also have been helped by the norms of discovery of DNA evidence as set forth in the ABA standards, which would have allowed expert review of the data in its raw form.  Without any discovery of the exculpatory DNA evidence, the case might have played out much differently, especially if Mr. Nifong had been able to escape censure.  If Mr. Nifong had brought the case to trial, it is difficult to predict how a jury would weigh the fingernail DNA evidence against potential alibi witnesses for Mr. Evans (Mr. Finnerty and Mr. Seligman had electronic alibis).  Juries have occasionally ignored strong alibi evidence in favor of weak or questionable evidence, as in the cases of Jonathan Fleming and Russ Faria.  Fortunately, the state's investigators came to the conclusion that Mr. Nifong should have, and Attorney general Roy Cooper ended the matter.

Update 5/18/2015
When Brad Bannon was preparing to defend David Evans, he read John Butler's textbook on DNA profiling.  His cross-examination of Dr. Brian Meehan was one of the pivotal moments of the case.  For the newest edition of his three-volume textbook Dr. Butler asked Mr. Bannon to comment on aspects of how a defense attorney would challenge DNA evidence.  Mr. Bannon replied in part, "Did the lab follow acceptable standards of DNA analysis?  Did it follow its own protocols?  Is the lab applying those standards and protocols consistently or selectively?  For example, why do you call a peak below 150 RFU as a true allele for one purpose, or in one case, but not for another?  If there are such internal inconsistencies, do they usually inure to the benefit of one side's theory of the case?  If so, is that evidence of bias?"

Mr. Bannon's point about consistency of threshold values for peaks is similar to one found in the textbook An Introduction to Forensic DNA analysis, 2nd ed. (Rudin, N. and Inman, K., CRC Press 2002, p. 121) states, “It is important to have some predetermined limit to distinguish what is signal and what is noise.”  Without a clear guideline, a scientist may make choices that benefit his or her preferred hypothesis, even subconsciously.

A partial list of authors and references

 William Anderson “Duke:  Why the DNA mattered.”  William Anderson teaches economics at Frostburg State University in Maryland.  He was one of the first bloggers to call attention to the many problems in the Duke lacrosse case.

William Cohan, The Price of Silence, Scribner, 2014.  Mr. Cohan is a writer on business affairs.

Paul Giannelli, “DNA Profiling” in Race to Injustice (2009), Michael Siegel, ed.

Paul Giannelli is Albert J. Weatherhead III and Richard W. Weatherhead Professor of Law at Case Western University in Ohio.  He was the Reporter for the American Bar Association Criminal Justice Standards on DNA evidence, and he specializes in the area of scientific evidence.  William Thompson explained, "The Reporter for an ABA Standards Group or Study Committee is the person whose job it is to take down and 'report' the findings, conclusions and determinations of the group or committee.  It is the reporter who prepares the first draft of the report and of the annotations and notes that accompany a report.  The Reporter usually works under the direction of a committee chair, but has independent responsibility for getting details right."

Peter Gill, Misleading DNA Evidence, Academic Press, 2014.  Peter Gill is Professor of Forensic Genetics at The University of Oslo, Norway and is a professor at the Norwegian Institute of Public Health.  He is chair of the DNA commission of the International Society of Forensic Genetics, and he has published more than 180 peer-reviewed papers.

Georgina Meakin and Alan Jamieson, “DNA transfer: Review and implications for casework,” Forensic Science International: Genetics 7 (2013) 434–443.  Professor Jamieson is the head of the Forensic Institute in Glasgow, Scotland.

Robert P. Mosteller, Exculpatory Evidence, Ethics, and the Road to the Disbarment of Mike Nifong: The Critical Importance of Full Open-File Discovery, 15 George Mason Law Review 257-318 (2008).  Robert Mosteller is J. Dickson Phillips Distinguished Professor of Law at The University of North Carolina, Chapel Hill.  He teaches Evidence, Criminal Procedure Investigation, Constitutional Criminal Procedure and co-directs the Trial Advocacy Program.

Joe Minor, “Touch DNA: From the Crime Scene to the Crime Laboratory,” Forensic Magazine 4/12/13.  Joe Minor teaches forensic science and is a forensic DNA consultant.

Suzanna Ryan “Touch DNA Analysis: Using The Literature To Help Answer Some Common Questions.”  Forensic Magazine.  Suzanna Ryan is a contributor to law enforcement and forensic magazines, and she is a consultant and expert witness.

Stuart Taylor and KC Johnson, Until Proven Innocent, St. Martin’s Press, 2007.

KC Johnson hosted Durham in Wonderland, a blog that was a major source of information related to the case.  Stuart Taylor is a freelance journalist and a contributing editor for the National Journal, specializing in law.  He is a nonresident fellow of the Brookings Institution.

William Thompson, Painting the target around the matching profile: the Texas sharpshooter fallacy in forensic DNA interpretation.” Law, Probability and Risk (2009) 8, 257-276.

William Thompson, “Tarnish on the Gold Standard.”  William Thompson is a professor in the Department of Criminology, Law and Society at the University of California, Irvine.  His areas of expertise include forensics and human judgment and decision-making.

Van Oorshot R.A., Ballantyne, K.N., and Mitchell, R.J.  “Forensic trace DNA:  a review.  Investigative Genetics 1 (2010) 1:14. doi: 10.1186/2041-2223-1-14.  Roland Van Oorshot is a member of the Forensic Services Department of the Victoria Police, Victoria, Australia.

Kenneth Williams “An Examination of the District Attorney’s Alleged Unethical Conduct,” in Race to Injustice (2009), Michael Siegel, ed.  Kenneth Williams is a member of the faculty of the South Texas College of Law in Houston, TX.

Gary Leiterman and DNA contamination

Jane Mixer was initially thought to be the victim of a possible serial killer, John Norman Collins.  Jane Mixer was murdered in 1969 near Ann Arbor, MI, but the DNA testing was done until 2002, a gap of approximately 33 years.  One item of evidence from the Mixer case showed DNA from John Ruelas, and several locations (on some panty hose and a towel) showed DNA from Gary Leiterman.  Despite the fact that the state had no other evidence against Leiterman, he was convicted of her murder in 2005.  John Ruelas was never charged.

A drop of blood from Mixer’s hand was preserved, and DNA from John Ruelas was found.  The prosecutor believed that the blood was from John Ruelas, but he did not offer a satisfactory explanation of how his blood came to be there.  Because Ruelas was four years old at the time of the murder and lived about forty miles away, the possibility that his DNA arrived instead via contamination must be considered.  As defense expert witness Theodore Kessis wrote, “The unexpected and never accounted for finding of John Ruelas’ profile on evidence in the Mixer case clearly demonstrates the proposition that contamination can and does occur between samples from different cases.”

The presence of Ruelas’ DNA is most easily explained by the fact that samples from John Ruelas were also processed in the crime laboratory of the State Police of Michigan contemporaneously with the samples from the Mixer murder.  But can the absence of Mixer’s DNA in the blood drop also be explained?  In 2004 C. Peel and P. Gill (“Attribution of DNA profiles to body fluid stains,” International Congress Series 1261, pp. 53-55) performed a series of experiments, in which a good DNA shedder handled the substrate (cotton or glass) for a blood stain, either before or after the blood from a different individual was placed on the substrate.  The blood was either diluted or the stain had been left to sit for some months, allowing for possible DNA degradation over time.  They used leucomalachite green as a presumptive test for blood.  Peel and Gill wrote, “…the more dilute or degraded the stain, the higher the contribution of the substrate handler’s DNA to the resulting profile.  A positive presumptive test could be obtained from samples when a profile originating from the body fluid was no longer detectable.”  Although it is tempting to associate the DNA that one finds in a stain to that stain, such an association is occasionally in error.  In some cases substrate controls can be helpful in determining whether or not DNA is associated with a particular stain.

It is essentially 100% certain that Ruelas’s DNA arrived on the items from the Mixer case via contamination, although the exact route is unclear.   Gary Leiterman’s DNA was also in the laboratory at that time.  No body fluid could be associated with Leiterman’s DNA associated with evidence from the Mixer case.  Sub-source DNA such as this is weaker in probative value than DNA associated with a particular tissue or body fluid.  Taking these facts and ideas into consideration, Leiterman’s DNA probably also arrived via contamination. 

A second line of evidence also implies that contamination is the most likely explanation.  A lab worker performed a negative control experiment during the time that a sample from the panty hose was being tested.  The negative controls will only show the presence of DNA in an electropherogram if DNA has been introduced unexpectedly into the experiment.  Theodore Kessis wrote, “Review of the electropherograms associated with this negative control sample (NEG 041902) reveals that it was contaminated, a fact that cannot be disputed since Dr. Milligan himself labeled it with a note indicated as much (Appendix 8 – Electropherogram sample NEG 041902).  Remarkably, Dr. Milligan stated in his 7/15/02 testimony that no contamination events had occurred during the course of his testing and that if any had, he would have documented them in his reporters (p. 141-21 and 142-4).  Equally difficult to rectify here is the fact that when asked if he had ever committed an error, Dr. Milligan’s replied that he could never recall making one.”  CBS News reported that, “Lab supervisor Jeffrey Nye says he retraced every step and he does not believe there is any issue of contamination. ‘No issue whatsoever,’ he says.”  This is an astonishing statement.

This lack of disclosure of a contamination event and a similar occurrence in the Adam Scott case (Peter Gill, Misleading DNA Evidence, Academic Press, 2014, p. 22) demonstrates that one cannot implicitly rely upon a laboratory to report accurately the results of negative control reactions.  Yet this is valuable information; a jury might choose to discount testimony from a lab if it knew contamination had happened.  It is also worth recalling what William Thompson noted in “Tarnish on the Gold Standard,” which is that the some laboratory workers tamper with negative controls in various ways.  For these reasons full disclosure of the negative controls in the form of raw (meaning unprocessed) data is the best course of action for a judicial system.  Students of the Knox/Sollecito case will not be surprised to learn that the position of the Michigan State Police crime laboratory was foursquare against disclosure of the raw data.  Professor Thompson wrote, “The Deputy Director of the Michigan State Police issued a statement on May 12, 2005 opposing ‘the allowance of releasing raw electronic data for subsequent manipulations using software and parameters not validated by the Michigan State Police Forensic Laboratory’ and declaring that ‘it is the position of the Michigan State Police Forensic Science Division that any release of this (sic) data for processing with non-validated parameters is tantamount to evidence tampering.’”²⁶  This is a self-evidently nonsensical position, and it also forces one to ask why the processing parameters chosen by a forensic laboratory are necessarily the optimal ones.

Conclusions

The DNA evidence against Gary Leiterman is compromised so completely by the presence of the DNA from John Ruelas that it scarcely should be called evidence at all.  Exactly how their DNA came to be on items of evidence from the murder of Jane Mixer is not known.  Dr. Theodore Kessis commented on the Benjamin LaGuer case:  “It is highly improbable that any given forensic DNA laboratory will take it upon itself to contact its accrediting bodies or the press and state for the record how often they make mistakes… To best understand the weaknesses associated with DNA testing we must rely upon the empirical, the occasions in which such deficiencies are revealed either by the press or internal review of a lab’s documentation of such problems by a defense expert.  A close look at either reveals that indeed many instances of DNA testing errors have lead to the false conviction of individuals.”  Regrettably, even the sworn testimony of laboratory personnel may be seriously in error, as in the Leiterman case.  Nor can a jury be counted upon to accurately weigh the odds of contamination.    Quite the contrary, juries sometimes discount alibi evidence, such as the Farah Jama case, or the fact that the defendant lived in one city and claimed never having been to the city where the crime occurred, as happened in the Adam Scott case.  Both cases are now generally believed to be instances of DNA contamination.

The Independent DNA Experts and the Electronic Data Files

Part 28 in the Knox/Sollecito case

Update, 13 June 2011

In the story “Knox appeal: DNA experts to request more time” from the AFP on 20 May 2011, Knox lawyer Carlo Dalla Vedova said “The experts asked the forensic police to hand over information essential to their report on the DNA. They still haven't received it and will therefore request a 40 days extension.” He added, “It's not the first time we've asked for the police to hand over this information,” He also said, “But they need the raw data they have asked for from the police to do so. We first asked for it in 2009 and it's still not been handed over.” This ends the debate about whether or not the forensic files were ever released to the defense during the trial of the first instance.
____________________

Judge Hellmann appointed two independent experts to review the DNA forensic evidence in Amanda Knox’s and Raffaele Sollecito’s appeal. Recently, the experts asked for more time, and reports suggested that they did not yet have access to documents the felt were necessary to carry out this task.

According to Candace Dempsey, forensic scientist under whose supervision the tests were carried out, Dr. Patrizia Stefanoni, turned aside this request. She wrote to Judge Hellman, “In reference to the request of acquisition of CD RAW DATA, one is obligated to explain that the information in the form of this file in the sequencer is never an integral part of the technical report, as far as the object being tested by the forensic geneticist, namely the DNA profile, and that it is already reported in the electropherogram printout, connected to the technical report on which all of the useful date and an evaluation of the genetic profile are reported… Finally, the request asked for by the expert consultants relative to the acquisition of the CD RAW DATA appears incomplete in so much as the name of the ‘sample file’ requested was not specified…”

To help me consider Dr. Stefanoni's refusal refusal, I have consulted with DNA forensics professionals Dan Krane and Jason Gilder of Forensic Bioinformatics, and I gratefully acknowledge their help. The continued lack of file release with respect to the DNA profiling of this case has been a recurring theme of this blog.

Her arguments against releasing further information are essentially:
(1) All of the necessary data are already in the paper printouts of the electropherograms.
(2) The request for data files is insufficiently specific.

Let us examine point (1) first. Dr. Stefanoni’s position appears to be the same as it was when Dr. Pascali was refused data, as noted in Raffaele’s appeal. Yet some of the electropherograms only provide the number of repeats, not the peak height for each peak. Peak heights are essential to evaluate peak height imbalance within a locus, which bears on the question of whether or not a sample is in the low-template range, and whether two peaks within a locus belong to the same or to two different individuals. Peak heights can also be used to quantify the severity of degradation when one compares DNA fragments of different lengths. Peak height ratios also help one to decide whether or not a small peak is a type of artifact known as a stutter. A careful examination of these small peaks is especially important in helping to judge what other DNA is present on the bra clasp besides Meredith’s and presumably Raffaele’s.

In addition, having the electronic data files allows one to calculate a run-specific limit of detection (Gilder et al., J. Forensic Science, January 2007, 52 (1), 97). This process sets a lower limit on the size of which peaks to accept, based on the amount of noise.

It can also be helpful in detecting a type of artifact known as a pull-up. There are four types of dyes used in DNA profiling, each with a different wavelength (color) of detection. Each dye is ordinarily detected in its own channel. Sometimes a large peak gives a small spurious signal because of bleeding from one channel into another (Butler, Forensic DNA Typing (2005), pp. 336-337; 384). According to Christine Funk and Dr. Simon Ford, “Pull-up can usually be identified through careful analysis of the position of peaks across the color spectrum, but there is a danger that pull-up will go unrecognized, particularly when the result it produces is consistent with what the analyst expected or wanted to find.”

Dan Krane was asked to give his opinion about the release of such files in a separate legal matter. He wrote, “I believe that a defense expert cannot competently evaluate the results of an STR DNA test without having access to the test’s underlying electronic data. In my experience, review of electronic data has often led directly to the discovery of important problems or limitations in the STR testing, or to alternative theories of the evidence, that would not have been apparent based on a review of laboratory reports or other laboratory records… In my opinion, review of the electronic data is as important as review of the laboratory’s written notes…There is no legitimate reason for a laboratory to refuse a defendant’s request to examine the electronic data.” (bolding mine) Finally, this blog has previously noted that the ABA standards explicitly call for release of the electronic data files.

Point (2) is equally difficult to comprehend. Clearly Dr. Stefanoni understands that the electronic data files are being requested, yet apparently wants specific file names. It is difficult to see how the independent scientists would know the file naming convention used in Dr. Stefanoni’s lab. Who does Dr. Stefanoni think can provide the specific file names?

Forensic Bioinformatics has a 10-point standard discovery motion, and point 6 covers files. The material should include:
(6.1) All collection files (such as injection lists and log files for an ABI 310 analysis).
(6.2) All GeneScan® files, including sample files and project files.
(6.3) All Genotyper® files, including templates/macros (see Request 5).
(6.4) All GeneMapper® files, including sample files (.fsa files) and project files (.ser files).
(6.5) If the data you are providing includes files from another case that are not pertinent to the instant case (e.g., sample files from another case included in the same run folder), then please identify those non-pertinent samples by name and laboratory code.

Clearly it is the job of the laboratory that performed the test to provide the file names.

Concluding remarks
The failure of Dr. Stefanoni’s laboratory to provide the data to the independent forensic scientists is a continuation of her refusal to provide them to the defense. There is absolutely no legitimate reason for her to do so. As Dan Krane noted, “It is a fundamental tenet of science that two reasonable experts should be able to independently arrive at the same conclusions after reviewing the same experimental data.”

Raffaele Sollecito's Appeal

Part XVIII in the Knox/Sollecito case

Update, 7/12/10

Here is a better translation of a portion of Sollecito’s appeal dealing with the bra clasp, provided by an anonymous translator:

“But there is more. The lack of full discovery was also found during the course of the trial:
- on July 18 2009, during cross-examination of the technical consultant of Raffaele Sollecito’s defence, Prof. Adriano Tagliabracci, there was an astonishing coup de theatre, in which the prosecution formulated a question which generated the suspicion that the Public Minister was in possession of further laboratory data never made available to the defence. In particular, the data relative to the quantity of the extract used for genetic analysis of the biological material found on the bra clasp of the victim;
- confronted with the reaction of the defence, the office of the Public Minister had to admit that there existed documents compiled by the scientific police never deposited at the conclusion of the investigations (contravening in this way the provision of Article 415 bis c.p.p.), nor sent to the Gup with the request for trial (in violation of Article 416, paragraph 2, c.p.p.).
In other words, the office of the PM denied a complete discovery of the documents of the investigation, extending this breach in the rights of the defence into the trial!”

This passage indicates that the prosecution failed to provide the defense with information relating to the single most important piece of evidence against Raffaele Sollecito all through the pre-trial and well into the trial phase. What is unclear from the appeal document alone is whether or not all information was ultimately released. However, other documents indicate that the prosecution never released all of the information that the defense teams sought.
___________________________________________________

The following is Google-translated from Raffaele's appeal document (pp. 51-57) with some minor edits. Although it is hard to decipher the meaning in places, it gives a picture of the defense continually asking for information on the DNA testing all throughout the preliminary hearing and trial, and the prosecution continually refusing. I have previously written about the importance of the electronic data files to a proper defense and about the near universality of this disclosure. I am eager to hear the prosecution’s explanation of how their refusal to hand over log files and electronic data files can be anything other than a trampling upon the right of discovery:


“And there is no doubt that, in terms of 'prosecution, failure to discovery on one of the main aspects of the process has been a considerable advantage over the defense, since the case involved Raffaele Sollecito is building right on genetic testing. In fact, it is useful to make a historical reconstruction of the case at hand:
- Following the conclusion of preliminary investigations, the defense could review the technical report of forensic genetic investigations in the signature of Dr. Patrizia Stefanoni, to which were attached tables who reported only the number of alleles, without which they would specify any other information or scientific data (for clarity, we define each variant allele of a gene, the genotype of an individual gene has a relatively Kit alleles he is in possession);
- Given the extreme importance of this for proof on indication of Prof. Pascali expert defense Raffaele Reminder, on June 24, 2008 - pending the period of 20 days under Article. 415 bis -, was made an initial request to Prosecutor acquisition diagrams in electropherograms whereby the forensic results came in content Technical report on forensic genetics. In particular, this request is have specified the absolute necessity (in order of presentation defensive pleadings and petitions under Article. 415 bis cpp) to have the aforementioned documents available to the consultant of the could examine it in detail;
- However, against all evidence, the prosecutor dismissed the request was deemed inadmissible 'because the option in all.art.415 cpp a cover secured and only those acts ';
- To deal with a similar unjust refusal, the defense made on July 3, 2008, a further instance in which it sought to 'acquire forensic laboratories of the numerical values and RFU peaks on all findings, or alternatively to obtain police Scientific CDROM containing the raw data and peak RFU. It was asked, Moreover, in case it was not possible to obtain copies of these data, that the Prof. Pascali was allowed to go to the police service science for a spot inspection of the data and make it directly computerized copy;
- Even that request, however, was inexplicably rejected on 11 July 2008 with the anodyne statement that 'on the charts of Forensic there are already bars with generic value on fluorescence peaks.
The defense then returned to a preliminary hearing with a serious cognitive gap, due to lack of awareness of scientific data used by the police in their scientific investigations. As we shall see, this story had a further follow:
- Preliminary hearing of 16 September 2008, the defense reiterated request to acquire numerical values RFU and peaks at all findings or, alternatively, to acquire the CD-ROM containing the raw data RFU and peaks. The request filed at that meeting was based upon both the incompleteness detected in the diagrams attached to the report, and by the repeated requests made to the office of the C>P> to obtain the missing data;
- After this request, the Gup could not help but detect need 'for purposes of deciding' to understand 'fully (...) as procedure has been adopted for the analysis in question, and whether there further elaborations that constitute the necessary support conclusions reached. 'Consequently, the Gup has ordered the hearing in heard of Dr. Stefanoni calling 'to accompany their testimony with computer media containing data requested by the defense. "The same judge in that case was aware the extreme importance of data capture whereas the investigations living would receive 'the most important elements all.impianto accusatory;
- On September 25, 2008 in observance of the provision of Gup is was deposited by association. Zugarini on behalf of Dr. Stefanoni, the CD-Rom containing the requested data. In this regard like to highlight as the immediate acceptance of the instance in question proves unequivocally from the outset the importance of having the data available full scientific;
- Hearing on September 27, 2008, was then represented to the Gup need to obtain additional documentation because incomplete data, therefore, was asked to Court to acquire the service files (log files);
- However, the Judge, after seeking an opinion to Dr. Stefanoni has also rejected the application based on an uncritical adherence to its comments, coming to claim that "not be needed information contained in the logs (data nell.esperienza forensic are never identified, resulting standards approved by the parameters internationally and which appear to be in use in all Laboratories engaged in Forensic Genetics).
Such an established procedure is not, of course, acceptable under no point of view, since the unconditional adherence to observations of Dr. Stefanoni, has led to support the contention accusatory, without the technical-scientific comparison of an expert third and impartial.
And indeed, if it had consulted a person not bound by any constraint on the charge (ie a neutral), the parties would certainly have benefitted. And 'quite clear that to ask to review its person (Dr Stefanoni) which produced that result is, in fact, nonsense: can not say, claim of self-discipline science by the same technician who testified that result. Any specialist will of course be pushed, albeit inadvertently, to defend the validity of their work, even in front of a simple request for clarification, and justify - after all, understandably, by his point of view - that all data provided are sufficient.

Was evidence showed the need to provide all documentation scientific defense.
It 'important to note that log files are much scientific data relevant, if not indispensable, especially in situations (such as the case in point) where the peak height must be assessed together with other parameters (such as area) is also higher to determine the plausibility or, conversely, the unlikelihood attribution to a person's DNA. So it is true that during the preliminary hearing in court, the Dr. Stefanoni - at the request of the consultant that the Prosecution asked if it was important given the "peak area" – said "Normally not. "However, what is more surprising is that the very Dr. Stefanoni candidly admitted that it had no other experiences on (p. 68 transcription.'s preliminary hearing Oct. 4, 2009). The opinion of Dr. Stefanoni, in fact, emerged in all its fragility when the consultant has made the defense Reminder noted a recommendation of the International Society of Genetic Lawyers, which states that the method used to interpret traces of mixed DNA and consider l.area l.altezza of peaks "(p. 99 of those transcripts.)
The GUP at this point, he invited Dr. Stefanoni to be received by October 8, 'the data relating to the calculation of peak areas related to the finding 165 / B indicated by prof. Pascali in the third column, in his name, all.allegato referred to the report filed, and diagram on the second run on the sample elettroferografica 36B.
The epilogue of this long run-up occurred on October 8, 2008 with the filing by Dr. Stefanoni the CD-ROM containing the data dell.amplificato electrophoresis sample DNA called 165 / B (finding piece of cloth with hooks), and the CD-Rom containing: Sample data info for the two races dell.amplificato electrophoretic sample of DNA known as 36 / B (Finding knife) electropherograms of the second electrophoresis.
As can be seen from the picture just outlined, the defense did not have provision, at the conclusion of preliminary investigations, all acts of investigation as required by law, so much so that this uncertainty lasted at least for the preliminary hearing stage (and beyond, as you say).
For example, consider the simple analysis of the peak - technical data that was provided to the defense until after the hearing at the hearing Preliminary Dr. Stefanoni, damaging irreparably the rights of defense -revealed that the amount of DNA used for analysis was quite modest, which led the consultant to the defense clarified, that the interpretation given to the forensic graphics has minimally complied with the guidelines laid down by the International Society of Forensic Genetics.
But there is more. The lack of full discovery was also found in during the process:
- On July 18, 2009, during cross-examination technical consultant Defense Raffaele Sollecito, Professor Adriano Tagliabracci, there was an astonishing coup de théâtre, as the accusation made questions that have created the suspicion that the prosecutor had hold additional laboratory data ever made available to parties. It was, in particular, data on the amount of extract used for genetic analysis of biological material found on the hook of the bra of the victim;
- In front of the defense reaction, the Office of Public Prosecutor had to admit that there are acts performed by forensic ever deposited at the conclusion of the investigation (in contravention Thus the provision of art. 415 bis cpp) or sent to Gup with the request of trial (in violation of Article. 416, paragraph 2, cpp).
In other words, the Office of the PM or denied full discovery of documents investigation, pushing this lesion of the right of defense to the debate! We will now give an account of evaluations of the Court of Assizes of Perugia. The sentence I can refer about the motivations of the order of rejection "of All.audience 14/09/2009 defenses of the defendants objected l.avv n the interference with the right defense, arguing in this regard, which were verified the void. After hearing the other party seeking to disregard l.exception, this Court issued l.decree, which read out with which l.exception of nullity was rejected and was available to proceed further "(p.7. above).
Passing the order of 14 September 2009, the Assize Court noted as follows: 'respect of invalidity l.exception advanced and illustrated memory today filed for violation of the right of defense question exception is not admissible, l.activity sampling and comparison of traces biological object activity.exception the nature of prior technical investigation covered with the possibility dall.article 360 for parties to participate in and comment in this context is not apparent that there has been no lesion of the right of defense, for the nullity that would assessed in a case of a 415 dell.article knows nzione not provided by any specific provision or by a 415 Criminal Procedure Code that principle of mandatory nature of the void during preliminary dell.audience compare 1/10/2008 GUP measure included in documents filed today in support the plea of nullity that aspect has been addressed and its exception has not been repeated in terms of which all.article Code 491 Nell.audience of Criminal Procedure also before this Court on 18.7.2009 issue was not raised in terms of the void under the compare the transcripts of the hearing on 18.7.2009, with particular Refer to pages 102 et seq and 114 et seq therefore logical to conclude that l.aspect in all.exception now advances to integrate a nullity ta question But to sanction by the Supreme Court ruled on the matter for example compare Cassation Judgement second section 29573 of 2006 and Section Supreme number 21593 2009 its fifth exception was belatedly raised. As for the broken and also being the subject of advanced today than that penalty is justified in the presence of a omitted on filing all.entire investigations in this case l.omission h a covered part of the supporting documentation'”