DNA analysis under scrutiny
Posted June 14, 2018 6:51 a.m. EDT
ALBANY, N.Y. _ The New York State Police crime lab is analyzing DNA evidence using controversial techniques that have been characterized as junk science and have led to shakeups at other labs around the country.
A Times Union examination of records documenting the practices of the State Police's Forensic Investigation Center also found that the lab may be manipulating DNA analysis to match samples to known suspects in crimes, and that a national accreditation organization removed details of that allegation from an audit that was released in February.
The issue centers on mixtures of human DNA recovered from evidence such as weapons or a victim's clothing. Scientists at some labs, including the State Police's Forensic Investigation Center, use controversial subjective analysis to pinpoint the DNA of individuals who may have left their genetic footprint behind.
Questions about the State Police crime lab's application of the so-called "combined probability of inclusion," or CPI, were raised in a recent Orange County murder case. Lab records obtained by the Times Union show a State Police scientist changed her initial written findings that concluded the suspect's DNA was not on an ammunition magazine associated with one of the handguns used in the crime.
The records confirm that the scientist, Cheryl L. Strevell, issued a second report three weeks later, on Dec. 8, in which she discarded four of the unique genetic markers that had indicated the DNA belonged to someone other than the suspect, Omarrio H. Morrison.
The adjusted report, initially based on 16 genetic markers, or locations, was sent to the Newburgh City Police Department and used as evidence _ along with Strevell's testimony _ to help convict Morrison of second-degree murder at his trial, which ended June 1 in Goshen.
After initially finding Morrison's DNA was not included at all 16 genetic locations, Strevell's final report found that he was included using only 12 locations. Strevell reported the chance of a random person being in the mixture was one in 14.33 million.
Her adjusted findings came after she had apparently looked at the suspect's DNA profile, which multiple experts said is improper but may be common practice at the New York State Police crime lab.
"It's like painting the target after you've already shot the arrow," said Tiffany Roy, an attorney and DNA expert who exposed similar problems at the Broward County sheriff's crime lab in Florida.
James W. Winslow, Morrison's attorney, said that there was other evidence against his client, but that the DNA evidence was damaging. "There is this concept in DNA called 'suspect-centric analysis,'" Winslow said. "It seems to me, and to the other people that I've talked to, that they really should be doing these things in the blind and not knowing who the suspects are or their DNA. ... That's not the way it's done."
Roy, who examined the State Police lab's DNA evidence in the Morrison case for the defense team, said the scientist's report was "a complete misinterpretation of the profile. They pretty much left out information that was questionable ... and tried to include information that would meet their guidelines" to match the evidence to the suspect.
In February, ANAB, a non-governmental national accreditation organization, issued an audit that flagged the State Police lab's work in three cases, including the Morrison investigation.
An earlier version of the audit, which was done under the umbrella of the ANSI-ASQ National Accreditation Board, had characterized the State Police lab's CPI methods as inconsistent and improper, and noted its scientists have been "selectively" removing DNA "loci" from samples that did not support their interpretations.
The section removed from the audit also indicated State Police scientists were comparing evidence samples to the DNA of a suspect during the analysis, according to a person outside the State Police who was familiar with the audit.
That practice, which is akin to looking at an answer before completing the work, is considered improper and may lead to wrongly implicating suspects in crimes.
Similar allegations rocked a crime lab in Austin, Texas, where their outdated and subjective analysis methods came under fire in 2015.
Beau Duffy, a spokesman for the New York State Police, said CPI "is a valid statistical technique."
"In response to the audit, the lab took corrective action to update and clarify procedures, and provided staff with additional training, to the satisfaction of ANAB," Duffy said. "The (state forensic commission's) DNA subcommittee also reviewed the findings and corrective actions, and issued a recommendation to renew the accreditation."
There is no question in the DNA community that serious issues have been found when scientists use "interpretive analysis" to pinpoint the DNA of individuals in samples that are mixtures of multiple people.
In a 2015 report by the National Institute of Standards and Technology, more than 100 crime labs took part in a study in which they used the interpretive method to analyze DNA mixtures. The study asked the labs to identify DNA samples found on a ski mask. Although there were four known contributors, more than 70 of the labs _ spread across 46 states but not identified in the study _ also matched the DNA on the ski mask to a fifth person whose DNA had been provided, even though that person had never touched the mask.
It's unclear why the criticism of the New York State Police lab's interpretive methods was removed from the ANAB audit that was issued four months ago.
Nita Bolz, a manager at the Wausau Crime Lab in Wisconsin who oversaw the audit, did not respond to a request for comment.
Another scientist who took part in the State Police audit said that ANAB requires them to sign a confidentiality agreement that prohibits the auditors from discussing their work.
Problems similar to those described at the New York State Police crime lab led ANSI-ASQ to shut down a lab in Washington, D.C., three years ago after questions were raised by the U.S. attorney's office about the reliability of that lab's DNA work. The lab's DNA analysis was suspended for 10 months and slowly resumed under tighter controls.
Roy, formerly a chemist at the Massachusetts State Police DNA lab, challenged the accuracy of the work at the Broward County crime lab two years ago in a formal complaint that helped spur a broader investigation of the facility's work.
Last week, Roy filed a formal complaint with the New York State Commission on Forensic Science, which oversees the state's crime labs, about the issues raised in the ANAB audit, including the analysis in the Morrison murder case.
Her complaint identified what she said were "serious errors" in the conclusion reached in the Morrison case. She also challenged the explanation that Russell Gettig, an associate director at the State Police lab, gave to the commission at a meeting in March when a commission member questioned the ANAB audit findings.
Gettig said during the meeting that he had examined the three cases flagged in the audit, including the Morrison murder case, and that "there was nothing improper, or incorrect, that went out in the report in those cases."
Roy, in her complaint, said the method used by the State Police in those cases to identify the DNA contributors "is unreliable and unsupported by good science."
"I agree with the ANAB auditors and I disagree with Dr. Gettig," she wrote in the eight-page complaint.
Roy's complaint said the issues cited in the audit of the State Police lab "closely resemble the issues addressed in the audit of the D.C. lab, including several findings of validation deficiencies and improper case handling."
Ray A. Wickenheiser, director of the State Police crime laboratories since 2013, was not available for comment on Wednesday.
It's not the first time the State Police lab has been alleged to use suspect-centric DNA analysis. In 2016, three forensic scientists who were snared in an alleged cheating scandal at the State Police crime lab filed a federal lawsuit claiming they were wrongly targeted for termination, in part, because they questioned the validity of the lab's traditional DNA-testing program.
The scientists, whose lawsuits are pending in U.S. District Court in Albany, accused the agency of using "suspect-centric" analysis, and alleged an unknown number of criminal convictions may have relied on inaccurate DNA results. They accused State Police leaders of pushing the cheating allegations to justify their sudden abandonment of a plan to implement a computerized DNA program that could have been used to retest old cases.