- Crime Lab
- Crime Scene
- Death Penalty
- Digital Forensic Insider
- Digital Forensics
- Evidence Collection
- Expert Forensic Voices
- Forensic Anthropology
- Forensic Psychology
- Impression Evidence
- Medical Examiner
- Mobile Forensics
- Police Procedure
- Sexual Assault Investigations
- Witness Testimony
A Pennsylvania judge ruled that the maker of a DNA-mixture interpretation software does not have to reveal the source code of the program as part of a murder trial.
TrueAllele, the program made by Pittsburgh-based Cybergenetics, was the focus of a discovery motion by the defense attorneys of Michael Robinson, a man charged with murdering two men in Allegheny County in 2013.
“The source code is not material to the defendant’s ability to pursue a defense,” Judge Jill Rangos of the Pennsylvania Court of Common Pleas ruled Thursday.
A black bandanna found at the crime scene is the subject of the legal back-and-forth. A witness said the garment was worn by the shooter. A local crime laboratory found that the DNA came from three or more people – and concluded in early 2014 that it couldn’t be analyzed.
But TrueAllele could analyze the complex mixture – and concluded Robinson’s DNA was in there.
Robinson’s attorneys contend the defendant is being deprived of his Sixth-Amendment right to confront his accuser, because he cannot analyze the source code.
The Commonwealth of Pennsylvania and Cybergenetics both argued releasing the source code could potentially put the company out of business – or force the company to decline to act as a witness for the prosecution.
The software maker said it was a victory for scientific accuracy – and entrepreneurship, since the tool was the product of more than a decade of development.
“Reporting accurate match statistics on DNA mixtures promotes justice. Using unvalidated software that gives wrong answers does not,” said Mark Perlin, the chief executive and founder of Cybergenetics.
The latest ruling cites that TrueAllele has been used in a litany of other cases – the first being the murder trial of a Pennsylvania State Trooper, Kevin Foley. The 2009 conviction was based, in large part, on the software’s conclusion that the chances of Foley’s blood not being part of a mixture at the crime scene were an infinitesimal 1 in 189 billion.
DNA mixtures have been a controversial topic at within the forensic community for years. But the topic has reached critical mass in the last year.
Perlin spoke with Forensic Magazine at length in November, in which he described Combined Probability of Inclusion, or CPI, as a “random-number generator” that lacked scientific foundations. TrueAllele is the solution, since it removes the potential of human error that is introduced by some of the subjective decisions required in the CPI process.
In a statement to Forensic Magazine this week, Perlin said the latest court ruling is just the latest approval among more than 500 cases in which the software has been used.
“Judge Rangos concurs with other courts,” said Perlin. “Scientists test executable software program on real data; they do not read source code text. TrueAllele’s reliability has been established through 30 validation studies, seven of them published after independent peer review.”
TrueAllele has competition. STRmix is another software program which also analyzes mixtures and potentially sorts them out. STRmix’s use has been more limited in criminal courtrooms – but it was approved for use during a Michigan armed robbery trial in December.