Back to Newsroom
16-Dec-2016
Cybergenetics responds to letter from Hillary prosecutor
Pittsburgh, PA
On November 1, at a prosecutor's request, Dr. Mark Perlin of Cybergenetics emailed District Attorney William Fitzpatrick of Syracuse, New York for information about a defense expert. A postscript noted that other prosecutors wanted Dr. Perlin to speak about DNA issues at an NDAA forensic evidence program, which DA Fitzpatrick had offered to arrange two years ago.
Mr. Fitzpatrick is past President and current Chairman of the Board of the National District Attorneys Association (NDAA). He was the lead DNA prosecutor in the New York v. Nick Hillary case. Mr. Hillary was acquitted in September.
The Hillary case is important because it highlights long-standing subjectivity in DNA mixture interpretation – forensic software generally lets operators choose the data that goes into the program. In Hillary, one data choice gave inclusionary statistics, while another data choice excluded him. The prosecution's DNA was precluded from evidence. Dr. Perlin talked about these results at Legal Aid and PCAST.
On December 8, DA Fitzpatrick sent Dr. Perlin a heated email reply containing inaccurate statements. He referred to the acquitted defendant as a "sociopathic killer of a helpless 12-year-old boy." He copied the email to almost two hundred people at PCAST, NIST, crime labs, and many prosecutor offices.
On December 16, Cybergenetics responded to DA Fitzpatrick.
Response
Dear Bill,
If you'd rather not share your Jamieson materials, a simple "no" would have sufficed.
The random match probability (RMP) method is used for major (over 50%) contributors to a DNA mixture where there is no dropout, and is not a suitable method for an extreme minor (under 1%) contributor. The New York State (NYS) Police threshold of 50 relative fluorescence units (RFU) is for their manual review of DNA mixtures, and is not applicable to statistical computer interpretation of mixture data.
As applied in your case, the STRmix results were subjective, requiring an operator to choose their DNA data and software parameters. Different choices gave different results, and could either include or exclude defendant Nick Hillary.
The New Zealand ESR company develops STRmix. ESR's December 2015 STRmix report used a 50 RFU threshold, which discards data that excludes Hillary. Using that report's same parameters at a 30 RFU threshold, STRmix excludes Hillary. Your own Onondaga County crime lab could perform this two-contributor calculation to verify Hillary’s exclusion.
STRmix works well within its validated limits. ESR's own mid-trial validation showed that with extreme (under 1%) mixtures, their software could not distinguish between a true contributor and a true non-contributor. The U.S. Department of Commerce's National Institute of Standards and Technology (NIST) has supported using STRmix for mixtures containing up to 3 contributors, down to a 20% component.
However, NIST has commercial conflicts of interest regarding ESR and STRmix. ESR has had their chief STRmix developer and proponent embedded inside NIST for over two years. NIST and ESR personnel work together crafting regulations that govern STRmix and its competitors. NIST favors limited software that can help them secure funding to "solve" a nonexistent DNA mixture "problem."
That NIST conflict underlies President Obama's Council of Advisors on Science and Technology (PCAST) unsupported findings on DNA mixture interpretation methods. First, NIST advised PCAST to impose artificial limits on the use of validated software that actually works well beyond such limits. Then, PCAST recommended that NIST get over ten million dollars to start testing, because of these nonexistent limits. With this funding, NIST could exert unwarranted regulatory control over forensic evidence.
PCAST and NIST are working together to undermine reliable forensic evidence. Courtroom use of well-established science that goes beyond NIST’s artificial DNA "limits" is at risk. This government intrusion punishes victims, innocents and society, while benefitting the guilty and corrupt. Your National District Attorney Association (NDAA) has strongly condemned the PCAST report for such reasons. You should be extremely concerned about pseudo-scientific government interference in criminal justice.
In Onondaga County, PCAST affects your own crime lab. The combined probability of inclusion (CPI) match statistic for DNA mixtures is not accepted as valid science. Past DNA mixture analysis for over fifteen years was based on CPI. All your DNA convictions that used CPI are at risk, either because of unreported potentially exculpatory "inconclusive" results, or inaccurate inculpatory match statistics used at trial. You and your lab were aware of these issues in 2011 when apprised of CPI failure at NYS Forensic Science Commission meetings, and again when I spoke to your office in 2014.
Cybergenetics has helped you with accurate TrueAllele® analysis that overcomes the bias and limitations of other DNA mixture interpretation methods. I testified in your county in the People v. Lewis Swift homicide, and in the People v. Frank Thomas weapons trial. TrueAllele results were accepted by stipulation in your People v. Ron Meadow homicide.
Cybergenetics would be happy to review all your past DNA mixture cases, assisting conviction integrity by identifying potential issues. When you need reliable science that goes beyond the limitations of other methods, we are here to help. As always, all initial TrueAllele screening is done free of charge.
Cybergenetics motivation in Hillary was to help prevent a grave scientific injustice. The company worked pro bono on the case, conducting a quarter of a million dollars worth of service without pay. Our scientists examined the data from all hundred and fifty items of DNA evidence, not just one fingernail item.
Over a three-year period, Cybergenetics assisted the police, prosecution and defense in understanding the DNA evidence. TrueAllele doesn't take sides. Our Hillary involvement did not help business. The company generally breaks even, investing heavily in science and pro bono justice.
Judge Catena acquitted Nick Hillary. No physical evidence connected him to Garrett Phillip’s death. When a defendant is found innocent, the evidence does not support a finding of guilt. Yet the prosecution still wins. As inscribed on a Washington wall, "The United States wins its point whenever justice is done its citizens in the courts."
I appreciate your sincere passion for justice. Perhaps not all of your forensic and government contacts help your cause. Please watch my talk again, and think about it over the holiday. Let me know if you would like to talk further. Best wishes for a Merry Christmas and a Happy New Year.
Kind regards. – Mark