Soldier death defence lawyer casts doubt on DNA expert's evidence

A DNA expert who testified against two men accused of murdering two British soldiers is prepared to exaggerate and tell partial truths in order to promote his technique, a court was told today.

A defence lawyer claimed Dr Mark Perlin was so focused on his "mission" to get his forensic method accepted within the scientific community that he lacked the objectivity to acknowledge its reliability was not yet proved.

Dr Perlin's evidence strongly linked Colin Duffy and Brian Shivers to the getaway car used by the killers of English Sappers Mark Quinsey and Patrick Azimkar, who were gunned down outside a Northern Ireland army base two years ago.

The American expert's statistical computer-based method of analysing genetic samples is relatively new and has never once been admitted as evidence in a UK or Irish court, and only on a few occasions in the United States.

On the eighth day of the murder trial at Antrim Crown Court, the forensic specialist strenuously denied the claims put to him by Shivers' defence counsel Patrick O'Connor QC.

Dr Perlin insisted he was objective and stressed that while he may have employed dramatic touches when describing his method to lay audiences in order to make it more understandable, he had never once exaggerated when presenting evidence in the scientific and legal arenas.

Judge Mr Justice Anthony Hart, who is sitting without a jury, had been hearing Dr Perlin's evidence before making a ruling on a defence application for it to be excluded from his final deliberations.

Sappers Quinsey, 23, and Azimkar, 21, were shot dead by the Real IRA as they collected pizzas with comrades outside Massereene Army base in Antrim town in March 2009.

Duffy (aged 43) from Forest Glade in Lurgan, Co Armagh, and Shivers (aged 46) from Sperrin Mews in Magherafelt, Co Derry, deny two charges of murder and the attempted murder of six others - three soldiers, two pizza delivery drivers and a security guard.

Dr Perlin carried out tests on data from a seatbelt buckle and a mobile phone found inside the Vauxhall Cavalier getaway car, which was abandoned partially burnt-out on a country road just a few miles from the shootings.

He said that a DNA sample found on the buckle was 5.91 trillion times more likely to be Duffy's than someone else's while a sample from inside the phone was 6.01 billion times more likely to belong to Shivers than another person.

Dr Perlin's technology has emerged in the last two decades as an alternative to the long-established human review technique to extract individual profiles from mixed DNA. His True Allele casework system has been available in the commercial market for last two or three years.

On his third day in the witness box, the expert was subjected to an intensive cross-examination by Mr O'Connor.

Producing literature and speeches previously written by Dr Perlin - one in which he described traditional methods as "dumbing down" DNA analysis - the lawyer claimed the academic exaggerated the worth of his own system at the expense of others.

He claimed he also had not made clear to readers that not all validation tests on his work had been finished and that in the one case he was previously involved with in England his evidence had not been admitted by the judge.

"You are prepared to exaggerate and tell partial truths in your mission to get your product accepted," said Mr O'Connor.

But Dr Perlin denied the claim.

"I would say that in a scientific arena or legal arena where exactitude is required, I don't believe that's the case," he said.

The expert said he may have added drama to some of his other writings, but only in order to make it more understandable for a lay audience.

With Duffy and Shivers watching on from the dock, Mr O'Connor conceded that True Allele may well become an important DNA method in the future.

But he told Dr Perlin: "You are still very much in development and you have not established your system so that it's accepted broadly within the scientific community."

Dr Perlin insisted that among scientists who develop new forensic technology his method was accepted and had been proved to work through validation.

However, the lawyer continued: "You lack objectivity and are not prepared to consider any reasonable concerns about limits to your system."

The expert said that was not the case, and noted that he and his team had resisted the temptation to release True Allele to the market for years while they developed it through 25 different versions.

"We need to maintain an objectivity in order to develop and release technology that meets our standards," he said.

Earlier, Mr O'Connor focused in close detail on the tests the expert conducted on the sample from the mobile phone.

Dr Perlin ran more than 100 tests on nine separate genetic samples - three originals and two subsequent enhancements of each of those. He said his final "likelihood ratio" of 6.01 billion was a representative figure of those outcomes.

Mr O'Connor noted that on some samples the computer had come up with a much higher ratio and on others a much lower.

He put it to the witness that there was a significance difference in the range of figures produced.

The expert said, statistically speaking, the difference was not marked.

"What was reported on was the centre of the distribution of likelihood ratios that were observed," he said. Mr O'Connor claimed Dr Perlin had only reported on a tiny fraction of all the tests he had run.

The expert said that was the accepted practice in such scientific analysis and rejected the claim he had failed in his "duty of candour" by not presenting all the findings up front.

"The common practice is to report on a statistic that is representative instead of overwhelming with a lot of detail which is not asked for - that is the common practice we have come to know," he explained.

The lawyer also asked Dr Perlin why he had not asked the computer to run tests that factored in four potential contributors to the DNA sample when another expert - Dr Emma Watson - had claimed there were possibly four.

Dr Perlin explained that Dr Watson was looking at all the samples collectively while he was referring to the nine individually, none of which showed four contributors.

Dr Perlin is set to conclude his evidence on Tuesday via video link from Pittsburgh in the USA.

The trial adjourned to sit again on Monday.