By Abigail Goldman (contact)

Fri, Aug 29, 2008 (2 a.m.)

A 9-year-old Jane Doe was sexually assaulted in her Carlin home 14 years ago. The man charged with the crime, Troy Brown, was getting drunk with Jane’s mom at the Peacock Bar the night it happened. He claimed he had stumbled home wearing his own vomit and went to bed.

His fingerprints were never found at the scene, but his DNA was.

Well, maybe.

This old Nevada case is suddenly new again — not only because the 9th Circuit Court of Appeals gave state forensic scientists a spanking in May for their inaccurate presentation of the DNA evidence, but because it may be an example of things to come, not just in Nevada, but everywhere.

When all other evidence fails, when the fingerprints and bite marks and ballistics and even witness testimony disappoint, DNA, that spiral stepladder of molecules, stands alone as forensic science’s armor.

Recently, however, small chinks in the armor have been starting to show, casting pinholes of harsh light on a science people like to think is infallible.

Problems began seven years ago in Arizona when a crime lab analyst scanning her state’s DNA database discovered two felons with incredibly similar genetic profiles, so similar that the FBI estimates the odds at 1 in 113 billion. The felons weren’t related. They weren’t even the same race. One was black, the other white.

That analyst then found dozens of unthinkable matches, raising questions about the accuracy of the FBI statistics and prompting other states to search their databases.

When the FBI found out about the data mining, a Los Angeles Times investigation revealed, agents worked to stop the searches and told state officials that as punishment they would be prohibited from accessing the DNA database.

This news made waves. Experts have suggested the matches were really not surprising given the unique way they were extracted from the data, a process more complicated than is worth explaining. But the damage has been done.

Right or wrong, the gold standard, DNA, is starting to lose a little of its

14-karat luster.

David Faigman, a University of California, Hastings, law professor who specializes in scientific evidence, told the Times, “DNA is terrific and nobody doubts it, but because it is so powerful, any chinks in the armor ought to be made as salient and as clear as possible so jurors will not be overwhelmed by the seeming certainty of it.”

When Troy Brown went to court on the sexual assault charge, a Washoe County crime lab expert testified that DNA revealed a 99.99967 percent chance Brown was guilty. Brown went to prison.

As the appeals court noted, absent the expert’s testimony, “there would have been insufficient evidence to convict.”

But jurors were given an incorrect calculation about the probability that Brown was guilty. It involved a computation also too complicated to explain, but common enough to have a name you can look up: the prosecutor’s fallacy.

The calculation was further complicated by Brown’s brothers — two in Carlin and two in Utah. Five men shared similar genetic profiles. The child victim flipped-flopped between two of them. She was unable to identify Brown as her attacker.

Laurence Mueller, a University of California, Irvine, professor who specializes in DNA typing, testified that the likelihood the DNA came from one of Brown’s brothers wasn’t 1 in 6,500, as the state said, but perhaps as high as 1 in 66 — 100 times more likely.

Brown, guilty or not, is now another example of DNA uncertainty. And while the world reconsiders DNA, Brown remains behind bars waiting for prosecutors to try to prove him guilty again.