"A
judge presiding over a trial in the 2011 murder of a 12-year-old boy in
northern New York State said on Friday that the prosecution could not
use a cutting-edge and controversial method of analyzing DNA data,
dealing a blow to its efforts to convict a former soccer coach accused
of the crime. In late July, Judge Felix J. Catena of St. Lawrence County Court held a hearing on the admissibility of a computer program known as STRmix in the trial of Oral Nicholas Hillary, who is accused of strangling Garrett Phillips in the village of Potsdam. The
police have testified that there is little, if any, physical evidence
against Mr. Hillary, 42, who has maintained his innocence since
suspicion first fell on him shortly after the murder. Mr. Hillary, a
black Jamaican immigrant who dated Garrett’s mother until shortly before
the crime, has contended that he is being targeted in part because of
his race. Garrett was white, as is much of Potsdam and the surrounding
county. Earlier
this year, the prosecution team indicated that it would use STRmix to
analyze a minuscule scraping from one of Garrett’s fingernails, and it
later suggested that Mr. Hillary was one of eight people in the nation
who might match the sample. The defense rebutted
those assertions and challenged the reliability of the program, which
uses an algorithm to determine probabilities for usually unusable data. On
Friday, Judge Catena agreed with the defense, finding that the data
could not be used at trial, a finding hailed by Mr. Hillary’s lawyers. “We
believe it to be the correct decision because it recognized what
happened here was unreliable and improper,” said Norman Siegel, a
veteran civil rights lawyer who has recently joined the defense team.........Seeking
evidence, the prosecution in the Hillary case had the asked one of the
creators of STRmix, a scientist from New Zealand, John S. Buckleton, to
analyze and develop a statistic for a possible match. Dr. Buckleton
testified at the hearing in July, which was followed by another hearing
on Aug. 17. In his decision, Judge Catena said the state crime lab initially analyzed the DNA in the case and sent it to a New Zealand government institute
that helped develop STRmix. But because the state crime lab was not
authorized to generate data for STRmix analysis, the judge wrote, Dr.
Buckleton “was forced to pick and choose data from different ‘reliable
sources’ and input parameters into the program in such a way that he
believed the system would tolerate.” The
judge concluded that “the lack of internal validation by the New York
crime lab, as candidly admitted” by Dr. Buckleton, “precludes the use of
the STRmix results.”.........The use of new computer programs
to analyze mixed — and extremely minute — portions of DNA has allowed
the use of some samples that were previously considered inconclusive.
But there is still ample scientific debate
about the use of such programs in cases where only a few cells may have
been recovered, and indeed, in the Garrett Phillips case, the extremely
small amount of material available was part of the defense’s argument
against its use. In
a telephone interview on Friday, Mr. Hillary said that he was “happy
that the judge came to the right conclusion.” The trial, however, is
still scheduled to begin on Sept. 6 in Canton, the county seat."
http://www.nytimes.com/2016/08/27/nyregion/judge-rejects-dna-test-in-trial-over-garrett-phillipss-murder.html?_r=0