Back to Newsroom
8-Sep-2016
New York judge again precludes STRmix from Hillary trial
Canton, NY
The Honorable Judge Felix Catena again precluded prosecutors from using "any conclusion reached by use of the forensic software tool STRmix" in the case of New York v. Oral Nicholas Hillary. Hillary, 42, is on trial for the murder of Garrett Phillips, 12, of Potsdam, NY.
After a Frye admissibility hearing, on August 26 the court rejected STRmix as reliable DNA evidence. Today Judge Catena denied the prosecutors' "motion to renew and reargue admissibility of DNA evidence."
Cybergenetics TrueAllele® technology examined over a hundred items, and showed that Hillary was not connected to the DNA evidence in this case. Prosecutors later retained DNA expert Dr. John Buckleton, an employee of New Zealand’s ESR company.
Dr. Buckleton used his STRmix software, and a second method called random match probability (RMP), to analyze DNA data from the victim's fingernails. He found an association. Dr. Buckleton told the court his "recommendation is to validate" STRmix in-house, which was not done. He also said "RMP overstates the weight of evidence in this case."
The defense submitted a response letter to Judge Catena on August 24. In Exhibit B, Dr. Mark Perlin of Cybergenetics described how picking data gave a biased result. "Buckleton chose a threshold of 50 rfu. But the fingernail evidence contains potentially exculpatory evidence between 30 and 50 rfu. And STRmix is validated for using more peaks at 30 rfu, not fewer at the higher 50 rfu level."
The Exhibit continued, "In fact, running STRmix at a validated 30 rfu threshold would exclude Hillary. The fingernail evidence is exculpatory. STRmix proves that Hillary’s DNA is not present."
"Judge Catena made a wise decision about validation," said Dr. Perlin. "Some forensic software works well within narrow limits. Overstepping those validated boundaries can cause mischief." STRmix, like the FBI's Popstats software, lets analysts choose their input data. Biased data choices can give the wrong answer. An incorrect match statistic can prejudice a jury about DNA evidence.
Links
- Second DNA hearing denied, evidence and testimony permanently tossed from Hillary murder trial - Watertown Daily Times
- Judge rejects DNA test in trial over Garrett Phillips’s murder - NY Times
- Motion to renew and reargue admissibility of DNA evidence - Decision & Order
- Misinterpretation of DNA evidence - Exhibit B
- Overcoming bias in DNA mixture interpretation - Cybergenetics
- Failing to interpret DNA mixture evidence - Cybergenetics