Evolving Forensic Science and The Texas DNA Mixture Review Project

“As far as the laws of mathematics refer to reality, they are not certain, and as far as they are certain, they do not refer to reality.” – Albert Einstein

We are a society proud of our technology. Much of our advancement is in forensic science. Since antiquity we’ve created methods to scientifically determine the invisible. Vitruvius tells the story of a suspected fraud. A king decreed a crown be made entirely of gold, and assigned an appropriate weight for it. When the manufacturer produced the crown, the king suspected it to contain silver. Archimedes was tasked with detecting the theft, while not destroying the crown. Eurhka! he exclaimed, when he used water to determine the different mass for silver and gold. He exposed the fraud, and his law of buoyancy followed.

Perhaps one of the strongest modern examples of our pride in forensic science is exemplified by the media. We have movies and television shows dedicated to people in labs finding fibers, fluids and DNA. Huge mathematical statistics are used to link suspects to crimes, and shed light on the “facts.” But, as most lawyers know, the reality of cases often varies from television.

The discovery of DNA, and the methods to detect it, is perhaps one of the most significant advancements of our time. The science of DNA is powerful. It can unite families, tie fathers to children, and help amplify the truth. However, because this science and the calculations around it are performed by humans, it is susceptible to human error, and so can be dangerous.

One example of the power of DNA comes from a very old case. In the early 1900s, a four-year-old boy went missing at a swamp in Louisiana. Eight months later, a parade was thrown when the boy was found in Mississippi, with Walters, a traveling handyman. Walters insisted the boy was not the missing child, Bobby Dunbar. He claimed the boy’s name was Bruce, and his mother, Julia, lived in North Carolina. Walters explained that Julia was in dire straits and gave him permission to travel with Bruce. Law enforcement brought the boy to Louisiana to visit with Mrs. Dunbar, who purportedly recognized him as her missing son. Julia too was brought to examine the boy, and insisted he was her son, corroborating Walters’ story. Julia tried to convince the court that the child was hers, but she had no lawyer and no money. The boy was claimed by the Louisiana family who won him in court and raised him as Bobby Dunbar. Walters persisted that he was innocent, and in 1914 went to trial on kidnapping charges. About twenty witnesses supported Walters’ story, but, alas, he was convicted. Decades later, a DNA test helped confirm the truth (as featured on This American Life). The science of DNA did not exist at the time of trial. But, a DNA profile showed Walters was innocent. The boy had been taken from his mother and given to strangers. In the end, the DNA test, helped give Julia her son back, in the form of reuniting his descendants’ with their blood family. DNA science can be powerful.

Recently, in 2015, the FBI notified labs around the country that the population database it used (since 1999) to calculate match statistics in criminal cases, contained discrepancies. DNA profiles in criminal cases often lead to “expressions of probability,” a reference to the odds that the DNA profile would randomly appear in a certain population (divided by race). DNA lab reports will typically list this information, and if there is a subsequent trial, a lab analyst will testify regarding this mathematical probability. For example, an analyst might testify to her findings by saying, “the probability of selecting a person at random with this DNA profile is 1 in 555,000,000 for (insert specific racial group).” Ultimately, this evidence can serve as the death knell for a client on trial. Often, this evidence is material to someone’s conviction. For example, it might be the only direct evidence in a case, or the supporting evidence used to corroborate an accomplice’s testimony. In any case, these huge statistical numbers often leave juries satisfied as to the link between a suspect, DNA evidence and the crime scene.

If, for instance, semen is collected during a sexual assault kit, and a DNA profile is determined that appears at random in 1 of every 225,300,000 persons, and the suspect has the same profile, it would be futile to argue against the characterization that there was a “match.” While the numbers for sole contributor DNA evidence are not affected, this example highlights another problem which appeared in DNA mixture cases.

A DNA mixture case involves a sample with the combination of biological material from two or more contributors. For the past few years there have been changes in this mixture interpretation with guidance from the Scientific Working Group on DNA Analysis (“SWGDAM”). In some cases of recalculation, like in a Galveston case, the statistical numbers made big jumps. For example, from 1 in 1.4 billion to 1 in 38. In other words, the probability that the DNA sample belonged to someone other than the suspect was 1 in 1.4 billion, linking that person to the crime scene. However, upon retesting with the new protocol, CPI (Combined Probability of Inclusion) was now 1 in 38, meaning there was a 1 in 38 chance the DNA could belong to someone other than the suspect. Considering the high burden of proof in criminal prosecution, beyond a reasonable doubt, these varying numbers matter.

Enter the Texas DNA Mixture Review Project (“Project”), whose director is Bob Wicoff, and funded by the Texas Indigent Defense Commission. With the cooperation of the Texas District & County Attorneys Association, information for people serving sentences in cases involving DNA mixture evidence is being gathered. The project has been tasked with determining which cases might benefit from recalculation (not retesting) in cases that have been disposed of using old protocols. With the help of lawyers and assistant directors, Scott Ehlers, Betsy Stukes and this author, cases from across the State are currently being reviewed. Different labs across the country use different protocols, so the effects will vary. Additionally, even though shows like CSI portray DNA evidence as being the ultimate factor in a criminal case, in reality, DNA evidence may not be as crucial. DNA mixture evidence may come from touch DNA, such as, off the handle of a gun, or door, and may be periphery to other direct evidence.

Thus, this project is focused on reviewing DNA mixture cases, that is, cases with a DNA sample that has more than one (preferably more than two) contributors, and in which there was a CPI issued (generally issued in DNA mixture analysis cases). During this review, the materiality of the DNA mixture evidence is assessed. More information can be found on the Texas Forensic Science Commission website.

A client who is potentially affected by this evidence is notified, and consents to this review. Once the initial phase of the review is completed, the client is notified of the Project’s findings. If recalculation is requested, then the case progresses to the next level of review. In any event, the Project is proud to be tasked with this process, and anticipates the numbers of cases needing review to climb. As a result, the Project will be seeking volunteer lawyers. If you are interested, or know someone who is, please feel free to email the author.

About Victoria Erfesoglou

Victoria Erfesoglou is licensed to practice law in California and Texas. She practices criminal defense in Harris and contiguous counties. She is a proud graduate of Gideon's Promise and Harris County's FACT Program. Victoria is extremely please to serve as the Assistant Director of the Texas DNA Mixture Review Project. She is deeply committed to indigent defense and fighting the systemic failures of the criminal justice system. She has certification in the Greek language and degrees in Psychology and Philosophy.

Twitter: @VErfesoglou