by J.D. Schmidt

Modern criminal investigations, especially cold homicide cases, often rely on what is known as “touch-transfer” DNA to identify the perpetrator. But in recent years, developments in DNA research have shown that there is an increased risk of falsely linking an individual to a crime scene through its use. One study detailed in December 2015 as part of Forensic Science International’s Genetic Supplement Series called attention to the “increased chance of detecting the transfer of DNA from individuals not involved in the commissioning of an offence to scenes and/or exhibits of the cross transfer of DNA between scenes or items.”

For the last four decades, forensic investigators have utilized increasingly sophisticated techniques to build a record of DNA profiling as investigators’ ultimate tool for providing seemingly unassailable identification of the individual human beings whose bodies these types of tissue and fluid samples once belonged to. Yet against this vaunted record stands the problem of proving how and when the samples in question ended up in the place discovered.

Touch-transfer DNA has falsely implicated people in heinous crimes who have never had any direct interactions with victims or crime scenes, including toddlers and even the dead. Add this to the reality that most touch-transfer DNA analysis relies on fragmentary DNA, which is often a decaying and degraded collection of cells that may come from several individuals.

When we delve below the pop culture, surface-level perception of DNA analysis into the complications and contradictions highlighted by scientific studies and criminal cases alike, we find that the accuracy of identification can all too often be highly questionable. However, as demonstrated by decades of trial proceedings and the ongoing work of groups such as the Innocence Project, it can also be a powerful tool for exonerating the accused and overturning wrongful convictions. In fact, sometimes it is the only tool available to right those grievous injustices.

Types of DNA Transfer

“Touch DNA” refers to DNA, most often from skin cells, that is transferred directly and left behind when a person touches an object. This type of DNA deposit is known as a “primary transfer.” Along with DNA from bodily fluids and hairs, touch or primary transfer DNA is a crucial form of evidence that authorities can use to indicate the presence of an individual at a crime scene. This is especially true when other types of evidence are scarce or unavailable. At the same time, touch DNA evidence carries serious risks. These include the possibility of crime scene and post-collection contamination, as well as the difficulty of distinguishing between primary and secondary sources of DNA.

Scientists and technicians use the term “secondary transfer” to refer to situations in which DNA left behind by a touch of some sort is transferred from one surface to another. A handshake can lead to primary transfer; touching a bottle, cup, doorknob, or car door afterwards can lead to secondary transfer, as one person’s DNA is transferred from a second individual’s hand to the next surface they touch. Secondary transfer can potentially hinder or even negate the interpretation of DNA evidence, leading to confusing or even wildly inaccurate identifications as to who was actually present at a crime scene or involved in the commission of a crime.

“Tertiary transfer” is yet another method in which DNA is moved through multiple vectors. When someone touches a surface that has previously been touched by another person, in the process picking up that first person’s DNA (in the form of skin cells, saliva, sweat, or other bodily fluids), this secondary transfer can become tertiary. Or as an October 26, 2015, article in The New Republic described, “Tertiary helps signify that the DNA ended up in a place three steps removed from the initial contact with its owner.” While alleged to be statistically far less common, or at least less detectable, tertiary transfer has been shown to occur in scientific testing and is theorized to have been responsible for some of the most seemingly bizarre results returned by crime scene DNA analysis.

It should be clear by now that the problem with focusing on the mere presence of touch-transfer DNA at a crime scene or on an incriminating item of evidence found elsewhere is that it’s nearly impossible to determine whether it’s the result of a primary, secondary, or tertiary transfer. However, the method of transfer is of enormous importance and often the difference between accurately identifying the guilty party and falsely suspecting a completely innocent person.

Christopher Zoukis provided a stark example of this problem in the September 2018 issue of CLN in which he wrote:

In 2008, European authorities were hot on the trail of a highly prolific serial killer and burglar. The “Phantom of Heilbronn” robbed jewelers, burglarized caravans and murdered multiple people, including a law enforcement officer.

The Phantom left forensic evidence all over the continent. His DNA was found at 40 crime scenes in Germany, France and Austria. Police offered a large reward for his capture, and spent an estimated 16,000 hours working the case. But there was one small problem with the case.

He was a she, and she wasn’t a criminal mastermind. The Phantom of Heilbronn was an elderly Polish factory worker who unwittingly contaminated the forensic swabs she manufactured. While she worked, she transferred her DNA onto the swabs. Investigators later transferred her DNA to the scene of a slew of unrelated crimes.

The Phantom of Heilbronn debacle is a good example of the dangers of DNA contamination. Inside the Cell author Erin Murphy draws a distinction between DNA contamination and DNA transfer. Contamination, Murphy argues, can be avoided or limited by employing best practices. DNA transfer, however, is a much more insidious problem, because it is utterly unavoidable. In two minutes, Murphy says, the average person sheds enough skin cells to cover a football field.

DNA Analysis

As a 2016 article in the journal Science asserted: “Its accuracy has made DNA evidence virtually unassailable. A landmark report published by the National Research Council in 2009 dismissed most forensics as unproven folk-wisdom but singled out DNA as the one forensic science worthy of the name.”

“The discovery of the human genome was a watershed moment in the development of modern forensic investigation techniques,” Zoukis wrote. “With the advent of DNA testing, investigators have become proficient at narrowing down the sources of biological material to a statistical near-certainty. When blood, hair, or semen is found at the scene of a crime, forensic scientists can now establish a genetic profile that will exactly match only one person out of about a quadrillion.”

Anthony W. Accurso outlined the process of DNA analysis in the October 2021 issue of CLN as follows: “Genotyping is a science built on probabilities. Sequencing the entire genome (all of a person’s DNA) of every suspect would be prohibitively expensive and time consuming. As a shortcut, the FBI has identified 13 core locations (known as “STR,” for short tandem repeat markers or “loci”) on a person’s genome that, when sampled, are likely to enable unique identification. However, different ethnic groups can require different numbers of sample loci to establish uniqueness. Interpol uses ten loci when sampling persons from Great Britain and greater Europe, but only nine loci are generally used when processing samples from persons native to the Indian sub-continent.”

If “crime scene evidence is in a very small quantity, poorly preserved, or highly degraded,” such that only four genetic locations can be sampled, then according to a 2008 Nature.com article on DNA fingerprinting, the probability of uniqueness drops to “roughly 1 in 331,” Accurso wrote.

Recently developed techniques such as Polymerase Chain Reaction (“PCR”) testing allow for sometimes-highly accurate identifications to be made from minute, microscopic amounts of DNA-bearing material. Samples of a few dozen cells or fewer can now be utilized. However, investigators, prosecutors, and the public at large still tend to think of DNA in terms of the large-sample, “smoking gun” technology that criminology and popular media embraced in the 1990s and early 2000s, when DNA evidence was primarily derived from sources such as relatively large samples of blood and semen.

As Accurso pointed out, “things get infinitely more complicated (and less reliable) when the imperfect sample is from more than one person.” He asserted that public education would go a long way toward correcting the “mystique of the infallibility of DNA evidence in the courtroom.”

Touch-Transfer DNA on Trial

Touch-transfer DNA has featured prominently in a number of high-profile cases in recent years. One of the most notorious is that of Amanda Knox, the U.S. exchange student tried, convicted, and ultimately exonerated in the 2007 murder of her roommate, Meredith Kercher, in Perugia, Italy. Knox and her then-boyfriend were both convicted of Kercher’s murder in their initial 2009 trial, based primarily on touch-transfer DNA found on the victim’s bra strap and a knife found in the boyfriend’s apartment.

Trace amounts of what was determined to be the boyfriend’s DNA were found on the bra strap. Both Knox’s and the victim’s DNA were alleged to have been detected in trace amounts on the kitchen knife, which was determined by forensic experts to not match stab wounds on the victim’s body. A third defendant—whose DNA, bloody fingerprints, and other evidence associated with him was found in the victim’s room and on her body—was convicted in a separate trial.

The boyfriend’s DNA on the victim’s bra and the mixture of Knox’s and the victim’s DNA on the knife handle were easily attributable to touch and/or secondary transfer. Knox was known to have used the knife while cooking in her boyfriend’s kitchen. Moreover, the alleged finding of the victim’s DNA on the knife handle was considered highly questionable by international forensic DNA experts.

Writing in a May 13, 2015, article in New Scientist, Idaho Innocence Project Director Greg Hampikian, who testified for Knox’s defense, stated that “on one swab from the blade, a minuscule trace of DNA was detected, just once during many analyses. It had some that was consistent with the victim’s. This finding was never repeated, despite many attempts.”

In fact, Hampikian noted, “[T]he DNA on the blade came from so few molecules that analytical instruments were pushed to read below the level that the FBI, my lab, or anyone I knew would go. We asked the Italian lab to supply validation of such a sensitive measurement, but they never complied. Despite this, Knox was convicted. DNA experts in the U.S. spoke out, and a new study on the knife was then ordered in Italy. This failed to repeat the DNA finding,” and that failure was a significant factor in both Knox and her ex-boyfriend eventually being exonerated by Italy’s Supreme Court in 2015.

A similarly high-profile case involving touch-trace DNA is that of the 2022 University of Idaho student murders in Moscow, Idaho. Four students were stabbed to death in a shared apartment near the campus on the night of November 13, 2022. The person accused of this crime, Bryan Christopher Kohberger, has allegedly been tied to the scene by a variety of circumstantial evidence, including touch-transfer DNA that police claim was detected on the button of a leather knife sheath found near one of the victims. According to a January 5, 2023, story in USA Today, police “got a warrant to go through trash from Kohberger’s family home in Pennsylvania, and obtained DNA from the trash, which they were able to identify with a high level of probability as belonging to the biological father of the person whose DNA was on the knife sheath,” thus establishing an alleged link to the accused. Kohberger, who maintains his innocence, is scheduled to go on trial in August 2025.

During the early 2000s, a cold case involving the 1969 murder of law student Jane Mixer near Ann Arbor, Michigan, shined an early spotlight on troubling issues involving touch-transfer DNA. Mixer’s case had been lumped in with a string of other slayings known collectively as the “Michigan Murders” and generally attributed to convicted serial killer John Norman Collins. It was reopened in 2005 when investigators subjecting cold-case physical evidence to DNA testing got a hit. The match was allegedly from touch-transfer DNA theorized to be from perspiration on Mixer’s pantyhose, as well as a partial match from blood found on a towel placed under her head.

The person whose DNA purportedly fit the profile was retired nurse Gary Leiterman, who lived about 20 miles away from the crime scene at the time of the murder. Prosecutors and police were thrown a curveball, however, when DNA from a separate sample—a blood drop found on Mixer’s body—came back as a match with convicted murderer John Ruelas. The problem this presented was that Ruelas was only four and a half years old in 1969. This confusing contradiction gave Leiterman’s defense the opportunity to call into question the crime lab’s DNA testing processes, alleging post-collection contamination or some other form of malpractice. Despite this bizarre result and the questions it raised, the jury was not convinced that the lab’s analysts had erred in matching Leiterman’s DNA to the other samples, and he was convicted and sentenced, ultimately dying in prison in 2019.

In his August 15, 2018, CLN story, Christopher Zoukis detailed a similar case in which secondary touch-transfer DNA played a troubling role. In September 2009, Yale graduate student Annie Le was murdered in a secure laboratory facility on the university’s campus. Her body was found stuffed into a mechanical chase space behind a laboratory wall. DNA samples from her body and the scene inside the wall yielded two different profiles. One was that of a lab worker who was eventually tried and found guilty of her murder. The other profile, drawn from samples taken both inside the chase and on the victim’s body, including the waistband of her underwear, yielded a hit in criminal databases as a local convicted felon. Zoukis wrote, “Unsurprisingly, investigators looked at the convicted felon first. That turned out to be a literal dead end—the man had died two years before the murder.”

In 2015, The New Republic described the scene inside the cramped space: “[Y]ears earlier,” the offender who yielded a hit “had spent one long, hot summer building the very mechanical chase in which the victim was found.” Because the space was “closed from ordinary traffic or regular cleaning, coupled with the building’s strict temperature and environmental regulation (as a result of its role as a scientific lab),” a large quantity of DNA from the construction worker’s sweat was left behind. Due to this sweating and other conditions, the CLN article added, “[The offender] also was very likely to have been what scientists studying DNA transfer would call a ‘good’ shedder. Some people are naturally prodigious shedders of biological material. Those with flaky, sweaty, or diseased skin are thought to be good shedders.”

This case could have taken a radically different turn if the construction worker had still been alive. According to The New Republic, “Absent a good alibi—in this case, the irrefutable proof of his prior death—the worker might have ended up implicated in the crime. His familiarity with the space, along with his prior record, might have been used against him to prove that he had special knowledge of a good place to dispose of the body.” That potentially could have led prosecutors to pivot away from the laboratory worker—who was eventually charged, tried, and convicted for Le’s murder—and instead focus on the wrong individual.

Importantly, touch-transfer DNA truly is a double-edged sword. It can exonerate the wrongfully convicted, but it can also implicate the factually innocent. Unfortunately, many within the criminal justice system fail to appreciate the true nature of touch-transfer DNA and its limitations.

The plight of Lukis Anderson serves as one of the most compelling cautionary touch-transfer DNA cases ever. In 2012, Anderson, 25 at the time, was living on the streets of San Jose, California, struggling with alcoholism. He was familiar to many in law enforcement and healthcare in the area, known as a “frequent flier” at local hospitals due to his habit of public intoxication. When a wealthy local tech investor was murdered in an apparent home invasion just 10 miles away, investigators found the DNA of three people at the crime scene. Anderson’s DNA was on the victim’s fingernails. As Forbes reported, “Law enforcement crafted a theory of the case based on this evidence and Anderson’s lengthy criminal record, dangling the death penalty over Anderson’s head.”

Anderson, who was so intoxicated on the night in question that he struggled to remember where he had been or what he had done, began to doubt his own innocence, telling his public defender, “Maybe I did do it.” However, the public defender’s own investigator was able to determine that Anderson had spent the night in a local hospital being treated for intoxication at the time of the murder. The lead investigator on the case, who was determined to understand how the DNA evidence could possibly have pointed to a man who simply could not have been there, dug further into what happened. He discovered that the two paramedics who responded to the scene of the murder were the same ones who had picked Anderson up and brought him to the hospital hours earlier that night. DNA from Anderson was suspected to have been transferred to the victim’s hand via a pulse oximeter or possibly via the paramedics’ uniforms.

As Zoukis observed, “undoubtedly many detectives are not as open-minded and flexible in their thinking…. Had one of those detectives led the investigation in Anderson’s case, it could very well have had a tragic ending.”

Scientific Research

Puzzlingly, there’s an apparent lack of interest among researchers in the U.S. in studying touch-transfer DNA. Most of the cutting-edge research that has pushed this particular sector of the field forward and delved into its problems and contradictions has come from abroad—notably, Australia, Europe, and the U.K. In fact, writing in an October 14, 2019, article for CLN, Steve Horn noted that “[a]lthough it’s an area that has become increasingly prominent in trial and appellate court litigation … CLN could only find a single lab in the U.S. that has published a significant piece of scholarship on the topic” of secondary DNA transfer. Not much has changed since then.

In his groundbreaking 1997 Australian study of touch-transfer DNA, which was published in the journal Nature and titled “DNA Fingerprints from Fingerprints,” forensic scientist Roland van Oorschot both expanded and complicated the field of DNA analysis. He inadvertently challenged the basic assumption of the exchange principle, which was theorized by French scientist Edmond Locard that a criminal actor will always leave traces of his or her presence at a crime scene.

Van Oorschot and co-author Maxwell Jones described both how a successful DNA identification could be made from the minute amounts of DNA left by touch-transfer and how the very sensitivity of DNA testing technology opened up opportunities for error. “We show that an individual’s genetic profile can now also be generated from swabs taken from objects touched by hands, providing a new tool for crime scene investigations,” they wrote. They then warned: “Our findings also demonstrate the need for caution when handling exhibits and when interpreting results.”

The Forensic Institute in Glasgow, Scotland, reached a similar conclusion in its 2013 study “DNA transfer: review and implications for casework.” According to the study: “It is important to consider indirect transfer in the evaluation of trace DNA in casework…. The detection of a DNA profile upon a surface cannot be considered proof of contact. Research has demonstrated that the quantity of DNA recovered and the quality of DNA profiles obtained are complex issues dependent on many factors.” The study additionally stated: “While direct contact may be the most obvious conclusion, there is insufficient scientific data to establish the most likely mode of transfer in any specific instance. Furthermore, it has also been shown that DNA can be transferred during the forensic examination of items. It may therefore be difficult to rely on the locations of the finding of the DNA to inform on how DNA was deposited.”

In 2019, van Oorschot was the lead author on another paper on these issues published in Forensic Science International: Genetics. The paper, titled “DNA transfer in forensic science: A review,” served as a meta-analytical survey of high-profile literature on touch-transfer DNA issues to that point. As Horn noted, despite employing such a broad survey approach, none of the authors hailed from the U.S. They were all based in The Netherlands, Australia, or Great Britain.

Van Oorschot and his co-authors highlighted once again the problems involved in forensic analysis of touch-transfer DNA, both in terms of accurately interpreting potentially mixed and transferred DNA from crime scene evidentiary samples and in terms of the possibility of contamination via handling and analysis. “This review demonstrates that over the last few years we have become aware of several factors affecting [secondary DNA transfer], but much more research needs to be undertaken to understand the impact of the many variables, build the data necessary to determine probabilities of different profile type occurrences in different situations, and to improve the accuracy of the profile interpretation given the uniqueness of each case scenario to be considered,” they outlined. “As the number of potential scenarios in which [secondary DNA transfers] are to be contemplated is infinite, there will be reliance on extrapolating from research findings. The research thus needs to be of high quality, broad scope, and sufficient quantity.”

The paper described a concerning lack of sophistication in the industry in regard to what they termed “DNA transfer, persistence, prevalence and recovery” or DNA-TPPR, stating that “far more research is still required to better understand the variables impacting DNA-TPPR and to generate more accurate probability estimates of generating particular types of profiles in more casework relevant situations. This review explores means of achieving this. It also notes the need for all those interacting with an item of interest to have an awareness of DNA transfer possibilities post criminal activity, to limit the risk of contamination or loss of DNA.”

Christopher Zoukis declared in his September 2018 CLN article: “There has been an unconscionable lack of interest among forensic scientists to study secondary transfer. But there is little doubt that what van Oorschot discovered is accurate.” Zoukis then went on to detail one of the few independent U.S. academic studies that has actually been done on the subject, the results of which supported van Oorschot’s conclusions. This was a study conducted by Cynthia Cale and other researchers with the Human Biology Program at the University of Indiana titled “Could Secondary DNA Transfer Falsely Place Someone at the Scene of a Crime?” and published in the January 2016 issue of the Journal of Forensic Science.

Cale and her colleagues conducted an experiment in which a group of people divided into pairs exchanged two-minute-long handshakes—after which one person in each pair then handled multiple knives. Each knife was then tested, and DNA from the persons who handled the knives was detected in almost every case. Notably, testing also detected the DNA profile of the person who had not touched the knife 85 percent of the time. Furthermore, 20 percent of the time, the non-touching person came back as the primary, and in some cases, the only contributor of DNA.

Cale rightfully characterized the findings as “scary.” She added that “Analysts need to be aware that this can happen, and they need to be able to go into court and effectively present this evidence. They need to school the jury and the judge that there are other explanations for this DNA to be there.”

Writing in the journal Nature, Cale and her co-authors proclaimed, “We urgently need to review how DNA evidence is assessed, viewed and described. Everyone in the medico-legal community—forensic scientists and technicians, DNA analysts, potential jurors, judges and lawyers for both the prosecution and defense—must know and understand the potential for mistakes.” Does anyone genuinely believe that the foregoing constituencies are aware of the potential for mistakes and the various ways a person’s DNA can be deposited at a crime scene or on a piece of evidence via touch-transfer?

Beyond the direct results of the handshake-to-knife transfers, another finding Cale and her co-authors highlighted was the fact that no matter how much they tried to sanitize their experiment, unknown third-party DNA was also detected in the results, demonstrating the troubling reality of post-collection contamination associated with touch-transfer DNA. Cale’s experiment provided a huge red flag for the entire U.S. criminal justice system. It should have compelled the DNA analysis industry and the forensic science community to engage in deep broad-based review of how DNA is handled in criminal investigations and prosecutions. But it didn’t.

Steve Horn’s article titled “U.S. Government Lab Withheld Groundbreaking Study for 5 Years That Can Help Defendants Question the Reliability of Certain DNA Evidence,” which appeared in the March 2019 issue of CLN, mentions that U.S. government entities often have a vested interest in holding back information that would benefit criminal defendants. His article described the results of a series of experiments performed by the Applied Genetics Group of the National Institute of Standards and Technology (“NIST”), a unit of the U.S. Department of Commerce.

The NIST study was published on August 1, 2018, in the journal Forensic Science International: Genetics under the title “NIST interlaboratory studies involving DNA mixtures (MIX05 and MIX13): Variation observed and lessons learned.” It was based on a series of experiments conducted in 2005 and 2013. The experiments involved dozens of state and federal government labs—69 of them in the 2005 experiment and 108 in the 2013 experiment. The labs voluntarily enlisted to perform analysis on electronic files of DNA profiles derived from mock physical evidence taken from a variety of mock crime scenes, including sexual assaults, murders, and robberies. According to the NIST study, “The goal of the MIX05 and MIX13 studies was to examine sources of variability in interpretation rather than instrument sensitivity or amount of DNA being examined. Therefore, these studies involved sharing electronic files of DNA profiles with study participants rather than sharing of biological samples.” The study’s authors then compared and contrasted the results, looking for variations and patterns in how the labs interpreted their analysis of information on different mixed-DNA samples from these artificial scenarios.

The fabricated DNA mixtures were created with a range of different levels of genetic material available from various contributors including victims, suspects, “persons of interest,” and relatives of suspects. The labs were also provided with profiles of some, but not all, of the suspects and persons of interest in the various cases to compare the mixture samples with. Some case scenarios also included profiles of “decoy” non-contributors. The information was supposed to have been derived from DNA samples found on a variety of different types of crime scene evidence, ranging from mock rape kits to pieces of alleged physical evidence such as a suspected murder weapon found near a crime scene or a ski mask found in a trash can blocks away from a crime scene.

The ski mask scenario from the 2013 experiment provided the most telling example of the problems presented by different labs using a wide range of different software, statistical analysis protocols, and interpretive strategies when analyzing mixed touch-transfer DNA evidence. The ski mask was supposed to contain a mixture of DNA from four contributors, two of whom were alleged suspects identified by an informant. The labs were provided with information from buccal swabs taken from these suspects, as well as the profile of a fifth individual described as a “known accomplice” of these suspects whose DNA was not actually in the mixture and had never touched the ski mask.

Seventy-four of the 108 labs that participated in MIX13 got it wrong and included the “decoy” fifth person in their interpretation. According to the study, only seven out of the 108 laboratories arrived at the correct answer using the FBI’s methodology for testing DNA, known as CPI, or combined probability of inclusion.

The study’s lead author, John Butler, explained that this part of the study was specifically designed to determine whether modern DNA technology like that used in the participating labs can, indeed, place someone at the scene of a crime that he didn’t commit. In an interview with Forensic Magazine, Butler stated that, “The mixture itself was designed to not show too many alleles. People would be tricked into thinking there are only two or three people there, instead of the four people that were really there. The way that it was designed was on purpose, to kind of help people realize that CPI can falsely include people—that was its purpose. And it demonstrated that really nicely.” As for the study’s implications for the wider field of DNA forensics with regard to the interpretation of data from touch-transfer DNA, study co-author and ski mask experiment designer, Michael Coble, is quoted in a March 7, 2016, Science article as saying “It’s the Wild West out there. Too much is left to the analysts’ discretion.”

The NIST study’s eventual publication in 2018 is of great interest to criminal defendants, their attorneys, and supporters. Publication in a peer-reviewed journal is one of the non-dispositive and non-exhaustive factors in determining the relevance and reliability for the admissibility of expert witness testimony under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993). But for five years prior to the study’s publication in Forensic Science International, the authors’ findings in the MIX05 and MIX13 studies were presented only in PowerPoint slideshows at conferences, rendering the crucial information inadmissible under Daubert. Considering the implications of the errors and inconsistencies the study brought to light, the years-long delay between the experiments and the study’s publication in a peer-reviewed journal is telling and concerning for all whose cases were impacted by the prosecution’s reliance on this type of biological evidence in criminal proceedings.

In a September 21, 2018, New York Times op-ed by Greg Hampikian, a Boise State University forensic biology and forensic science professor, he criticized the disturbing gap in time between the NIST Applied Genetics group’s experiments and the study’s publication in a peer reviewed journal. “I first learned about the results of this study in 2014, at a talk by one of its authors. It was clear that crime labs were making mistakes, and I expected the results to be published quickly,” Hampikian wrote, “[b]ut years went by before the study was published, preventing lawyers from using the findings in court, and academics from citing the results in journal articles. If some of us had not complained publicly, it may not ever have been published.”

In a May 2018 Forensic Magazine interview, Hampikian described the consequences of the government’s choice to delay publishing the NIST study’s results in a peer-reviewed journal: “This is about five years of people being convicted by bad interpretations…. This is a huge story. It’s the problem with DNA—the one everyone trusts. If it was contaminated peanut butter, or faulty airplanes, or airbags that failed, wouldn’t NIST have felt compelled to do something more than just a couple PowerPoint shows? This is not just a case of salmonella—this is 20 years in prison. Some of these people died in prison.”

In addition to, and perhaps as a primary source of, their foot-dragging on the issue of peer-reviewed publication, the NIST study’s authors had conflicts of interest related to their employment with and studies at laboratories controlled by the FBI and other federal government entities. At the time of publication, two of the authors also sat on the board of Forensic Science International, which published the study, at the time of its publication. These conflicts of interest—particularly those related to law enforcement—may help explain why the authors inserted what Hampikian called “carve-out” language into the study’s publication, potentially derailing efforts to cite the study’s results in a Daubert hearing.

“Overall, Hampikian says he believes that the field of forensic science has yet to divorce itself from police academies and crime labs, which may go to explain the lack of in-depth focus on the issue of DNA transfer in the U.S,” Horn stated in his in-depth article in the October 2019 issue of CLN. “As Exhibit A, he pointed to the fact that NIST’s forensics science unit, headed up by Butler, has been put in charge of doing a federal governmental review of forensic science techniques as a response to recommendations made by the National Academy of Sciences in a 2009 report.” Such a review is badly needed. New techniques are in development that can potentially help with accuracy and dependability of touch-transfer DNA analysis, but more peer-reviewed, publicly accessible study is necessary at every level.

A recent Australian study investigated the possibility of secondary and tertiary DNA transfer via pets. As detailed in the March 2023 issue of CLN, the Victoria Police Forensic Services Department, forensic science researcher and Ph.D. candidate Heidi Monkman, and Dr. Mariya Goray of Australia’s Flinders University collected human DNA from 20 pet cats from multiple households. “A whopping 80% of the samples contained detectable levels of DNA. And ‘interpretable profiles that could be linked to a person of interest were generated from 70% of the cats tested,’” according to CLN.

Study co-author Goray summed up their results: “This type of data can help us understand the meaning of the DNA results obtained, especially if there is a match to a person of interest. Are these DNA findings the result of criminal activity or could they have been transferred and deposited at the scene via a pet?” For some people who find themselves the subject of a police investigation, the answer to that question could have life-altering implications.

Problem Exists Between Keyboard and Chair

Adding to the complexities of minute and fragmentary samples from multiple subjects potentially mixing together—and then being transferred across multiple vectors—is the reality of the imperfectly designed and operated tools used to decode them. Modern DNA analysis has come to rely on highly sophisticated computer software to interpret the patterns and deduce the probabilities involved in attempting to assign criminal responsibility. Unfortunately, as CLN and other outlets have documented in multiple articles in recent years, how this software is written and deployed and how its results are interpreted by forensic analysts and technicians often vary widely from lab to lab and organization to organization. Mixed samples are vulnerable to misinterpretation due to variations in the fine-tuning of the mathematics involved and the data input through the software, as well as the training and experience of the technicians interpreting the data it spits out. This variability can lead to potentially devastating impacts on both the accused and the victims or survivors of crimes.

Writing on issues in software-enhanced forensic DNA analysis in the August 2019 issue of CLN, Michael Berk stated that with contemporary technology “not only is it possible to run tests on very small samples, but the practice of using software to sift ‘irrelevant’ information from the materials under analysis, while retaining enough fidelity to hold up in court, has expanded DNA profiling far beyond its early boundaries.” Modern labs that do this high-tech work focus primarily on just forty short strands of DNA, or “alleles,” because these have been observed to be the sections of DNA that vary the most between individuals within human populations. “It’s extremely unlikely that two different people will share all 40 of these ‘genetic markers,’” according to Berk.

That said, a particular set of alleles showing up in final results does not necessarily mean that a person whose DNA has all of those alleles contributed to the sample being analyzed. Berk highlighted the following example: “There could have been two (or more) people, each having some of the indicated alleles, who supplied material for the sample. (Imagine two contributors, each having alleles 1, 2, 3, 4 and 5, 6, 7, 8, respectively—the results would include the set 3, 4, 5, 6 … but that doesn’t mean that someone having the distinct makeup 3, 4, 5, 6 contributed to the sample.)”

The type of computer program used in the interpretation of complex DNA mixtures is known as probabilistic genotyping software (“PGS”). These programs are built on statistical and biological modeling that is intended to determine the probability that the information revealed by testing indicates the presence of a specific individual’s DNA. The software is supposed to be designed to account for the types of anomalies regularly encountered in DNA analysis, and its calculations result in what is known as a “likelihood ratio.” The greater the likelihood ratio, the greater the probability that the test-sample DNA and the suspect’s DNA can be considered a match. “This number,” as Berk explained, “takes into consideration the frequency with which each allele (in partnership with other alleles) appears in the general population, and is an estimate of how likely it is that the particular set of alleles arising from the test of the sample came from the specified human subject.” Contrary to the popular conception of DNA always delivering absolute certainty, these probabilistic models can all too often lead to results that remain subject to interpretation.

Therefore, how the software these test processes rely on is designed and how its results are interpreted by technicians, scientists, detectives, prosecutors, judges, and juries can be critical for the outcomes of criminal trials that hinge on DNA evidence. On the one hand, computers are an absolutely essential tool for doing the complex calculations necessary for modern forensic DNA analysis. On the other hand, poorly written or deployed software from a variety of competing and proprietary vendors can lead to highly variable and conflicting results for public and private labs alike. In fact, public forensics labs in New York, Texas, and Washington have had to temporarily shut down testing or switch tools because of flaws discovered in DNA testing tools being used.

The training and professional practices of the technicians who conduct these tests and interpret their results for law enforcement and the courts can also be a problem. Differences in experts’ use of PGS may dramatically affect the results of DNA analyses. Different labs might produce different answers from the same sample simply based on variations in the programs used and their settings and controls, and such differences may call into question the reproducibility of test results.

Emanuel Fair’s Story

On June 11, 2019, a King County, Washington, jury found Emanuel Fair not guilty of first-degree murder. This ended an 11-year odyssey through the state justice system, nine years of which Fair spent incarcerated as one of the longest pretrial detainees in the state’s history. This tortuous journey was precipitated by a harrowing combination of prosecutorial overreach that Fair claims was motivated by racism and circumstantial evidence—of which touch-transfer DNA was the most important piece.

On Halloween night in 2008, Fair was the only Black person in attendance at a group party held by multiple tenants of a Redmond apartment complex. At least 50 partygoers, including residents of the complex and guests like Fair, moved freely between multiple rooms in multiple residents’ homes within the complex, particularly those of the party hosts/organizers. In the aftermath of the party, one of the hosts, 24-year-old software developer Arpana Jinaga, was found murdered in her bedroom. She had been sexually assaulted and strangled. Fair, who had spent the night at the apartment of one of the party’s other hosts, quickly became a suspect.

Fair told the police that he had been in and out of Jinaga’s apartment over the course of the night, hanging out in her bedroom with a group of other partygoers and at one point using her bathroom. Other partygoers corroborated his account. He also helped clean up towards the end of the party, including in Jinaga’s apartment. Investigators found DNA from at least three other men—including the victim’s ex-boyfriend and two residents of the apartment complex—in Jinaga’s apartment, on her body, and on items recovered from the complex’s dumpster.

Fair had a prior conviction for statutory rape on his record. Investigators, guided by King County Senior Deputy Prosecuting Attorney Jeff Baird, quickly closed in on him as their prime suspect. Their zeal to build a case against Fair led Redmond police and King County investigators to seemingly ignore evidence that potentially implicated others. They even provided immunity to one initial suspect, the victim’s next-door neighbor, in exchange for him providing evidence to bolster their case against Fair.

Despite their eagerness and myopic focus on Fair, the complexities of the case meant it was not until 2010 that authorities arrested him and officially charged him with sexually assaulting and murdering Jinaga. His family could not make bail. Over the course of a lengthy pretrial process and two trials—the first of which ended in a jury hung 11 to 1 in favor of acquittal and the second of which finally resulted in his acquittal—Fair would spend over nine years in the King County Jail. All of this happened because investigators pursued Fair with tunnel vision and where only able to build a flimsy circumstantial case against him that was based almost entirely on minute amounts of touch-transfer DNA evidence.

As reported by The Seattle Times on April 15, 2023, a U.S. District Judge backed up Fair’s allegation, saying that “evidence and arguments presented by attorneys for Emanuel Fair could support his claim that racism motivated police to target him for the rape and murder of Arpana Jinaga to the exclusion of other viable suspects, and that King County prosecutors were negligent in pursuing the case.”

The Times added that the judge’s ruling “points out, based on the pleadings, that police found DNA from three other men on key pieces of evidence, including the bootlace used to strangle her and on a motor oil can and bathrobe tied to the crime scene. The pleadings say police also found DNA from Jinaga’s white male neighbor, who was seen at the front door of her apartment just hours before she was killed, near her body.” The article specifically focused on the controversial touch-transfer DNA that the police and prosecution turned into the centerpiece of their case: “The complaint alleges that Fair’s DNA was in such minute quantities that it had to be sent to a special laboratory to identify it, and his complaint questions the validity of that conclusion.”

The company that performed this specialized computer analysis of the mixed, touch-transfer DNA in question, Cybergenetics, still has a post on its website touting the use of its product in Fair’s case, despite his acquittal. The page about Fair’s case includes a note that the Washington State Patrol Crime Laboratory “developed mixed DNA data from the evidence items” but “could not draw conclusions from the data due to its complexity.” It then touts the results of its TrueAllele computer DNA analysis produced for the prosecution. As of this writing, Emanuel Fair’s lawsuit against Redmond and King County authorities is ongoing.

Conclusion

Despite its growing use and importance in criminal investigations and prosecutions, a troubling number of law enforcement officers and prosecutors seem to lack a basic understanding of what touch-transfer DNA is and its critical limitations as standalone evidence. Attorney Marina Medvin wrote in a September 20, 2018, article for Forbes, “they expanded the search for touch-transfer DNA to all objects and surfaces, irrespective of the ability to find other identifying evidence connected to that DNA, such as a fingerprint. This led to the prosecution of individuals based on DNA from low-template and low-quality samples not connected to other identifying data. Moreover, prosecutors failed to distinguish the unique nature of touch-transfer DNA and the likelihood of random and innocent touch-transfer origins, presenting it to juries as the equivalent of a smoking gun.”

Consequently, according to Medvin, “when prosecutors present to a jury touch-transfer DNA evidence with the same oomph as large-sample DNA evidence, the jurors, under the influence of pre-set expectations for scientific evidence to prove culpability and the common notion that DNA evidence is inherently trustworthy, feel compelled to convict. The result is touch-transfer DNA can readily lead to conviction of the innocent.”

Because members of the general public remain largely ignorant of what touch-transfer DNA is, they conflate it with the now familiar large-sample DNA evidence, which truly is the forensic evidentiary gold standard. But the same cannot be said about touch-transfer DNA. With ever-smaller samples of genetic material used as evidence, the possibility of touch-transfer and contamination of samples with DNA from a wide variety of other sources is exponentially increased. Add to that the lack of clear and uniform standards with respect to software and hardware for touch-transfer DNA analysis, and the risk of wrongful convictions is exacerbated.

Given what we already know about touch-transfer DNA evidence’s ability to place a person at a location he’s never visited as well as place a person who’s been dead for years at a recent crime scene, the relative lack of new research into touch-transfer DNA and overall failure to properly educate police and prosecutors about the serious limitations of the evidentiary value of touch-transfer DNA as standalone evidence is alarming. The misuse of and overreliance on touch-transfer DNA evidence poses a high risk of wrongful convictions. CLN will continue to shine a light on this issue because there hasn’t been much improvement since CLN began covering it back in 2017. As Medvin concluded in her Forbes article, “As of now, anytime we touch a public surface, we remain fair game for criminal suspicion based on touch-transfer DNA.”  

Sources: sciencedirect.com, forbes.com, Criminal Legal News, investigationdiscovery.com, innocenceproject.org, unh.edu, nij.ojp.gov, forensicmag.com, science.org, seattletimes.com, genome.gov, lexology.com, usatoday.com, labmanager.com, newatlas.com, nature.com, discovermagazine.com, newscientist.com, cbsnews.com, newrepublic.com, cybgen.com, fsigenetics.com

Subscribe today

Already a subscriber? Login