F.B.I. Audit of Database That Indexes DNA Finds Errors in Profiles

An Article of Interest to All

N.Y. Times 

A version of this article appears in print on January 25, 2014, on page A15 of the New York edition with the headline: F.B.I. Audit of Database That Indexes DNA Finds Errors in Profiles.

F.B.I. Audit of Database That Indexes DNA Finds Errors in Profiles

By Joseph Goldstein JAN. 24, 2014

The Federal Bureau of Investigation, F.B.I ,  in a review of a national DNA database, has identified nearly 170 profiles that probably contain errors, some the result of handwriting mistakes or interpretation errors by lab technicians, while New York State authorities have turned up mistakes in DNA profiles in New York’s database.

The discoveries, submitted by the New York City medical examiner’s office to a state oversight panel, show that the capacity for human error is ever-present, even when it comes to the analysis of DNA evidence, which can take on an aura of infallibility in court, defense lawyers and scientists said.

The errors identified so far implicate only a tiny fraction of the total DNA profiles in the national database, which holds nearly 13 million profiles, more than 12 million from convicts and suspects, and an additional 527,000 from crime scenes. Still, the disclosure of scores of mistaken DNA profiles at once appears to be unprecedented, scientists said.

In some cases, the discovery of an error has enabled the authorities to identify new suspects in cold cases. One such discovery has breathed new life into the murder investigation of a man found bludgeoned to death in the Bronx in 1998. It also led to new matches in two rape cases in New York City in the 1990s, although the statute of limitations for prosecution appears to have expired. In these examples, the errors were found in the DNA profiles taken from the crime scenes rather than from people convicted of crimes.

The errors had the effect of obscuring clues, blinding investigators to connections among crime scenes and known offenders. It remains to be seen whether the new DNA evidence will cast doubt on any closed cases.

Dan Krane, a biology professor at Wright State University, said the disclosure was the government’s clearest acknowledgment to date that “there are mistakes in that database.”

The mistakes were discovered in July, when the F.B.I., using improved software, broadened the search parameters used to detect matches. The change, one F.B.I. scientist said, was like upgrading or refining “a spell-check.” In 166 instances, the new search found DNA profiles in the database that were almost identical but conflicted at a single point.

Alice R. Isenberg, the chief of the biometric analysis section of the F.B.I. Laboratory, said that most of the 166 cases probably resulted from interpretation errors by DNA analysts or typographical errors introduced when a lab worker uploaded the series of numbers denoting a person’s DNA profile.

“We were pleasantly surprised it was only 166,” Dr. Isenberg said, referring to the number of cases. “We were quite worried it would be much higher than that,” she said. “These are incredibly small numbers for the size of the database.”

It is possible that some of these cases might involve DNA from different people with remarkably similar genetic makeup. But in a number of cases, the slight differences have already been traced back to errors in the profiles stored in DNA databases: What appeared to be DNA profiles of different people was, in fact, the profile of the same person’s DNA, according to documents.

A review by the New York City medical examiner’s office determined that six of the 166 cases were attributable to errors in DNA profiles it generated from crime scenes in New York, according to a letter the office sent to the state panel.

Since the discovery of the 166 errors at the national level, the New York State Police has also changed the search parameters used to comb DNA profiles in the state database, finding additional errors. The state database includes profiles that do not meet the F.B.I.’s standards for inclusion in the federal database because they come from DNA that is too degraded or that came from complex mixtures of DNA at crime scenes.

In two of the errors discovered in the state database, analysts at the city’s medical examiner’s office made mistakes as they tried to discern a DNA profile from raw data, which appears graphically as a series of peaks, somewhat like the record a seismograph produces.

“These revelations spotlight how human error can detract from the reliability of the testing process,” said Alan Gardner, the head of Legal Aid’s DNA unit, which is challenging the city medical examiner’s methods for discerning DNA profiles in complex mixture cases.

In court, prosecutors often describe the strength of DNA evidence against a defendant with numbers that can run into the billions — expressing how unlikely it is that a person chosen at random would also have a DNA profile linked to the crime scene. But the rate of errors by a lab or a technician, a less dramatic topic, can be a much more relevant statistic, many defense lawyers and some scientists said.

“If we say there is a 1-in-10-quadrillion chance that someone else might have the same DNA profile, but there is also a 1-in-10,000 chance that there was a mistake in generating the profile, the only number the jurors should be paying attention to is the error rating, said Dr. Krane, who was once on a forensic science commission for the State of Virginia and now consults with defense lawyers on DNA cases.

Outside of New York, it is not clear which states have undertaken a review of errors. At the national level, there has not been a full accounting of the cases, which defense lawyers and some former laboratory scientists said was troubling. The lack of follow-up underscores the patchwork nature of the nation’s DNA databases. While the F.B.I. administers the software linking the state and federal DNA databases, it knows little about the actual profiles in the state databases.

“We don’t have oversight,” said Dr. Isenberg, of the F.B.I. Laboratory. “It’s up to the labs to take the information and act on it.”

In an interview, Barry C. Scheck, a co-founder of the Innocence Project, which seeks to overturn wrongful convictions, commended the F.B.I. for what he said was an important quality control initiative. “It’s hard for them to dictate what follow-up the state and local authorities will undertake,” he said.

But some former government scientists said the F.B.I. had to study the errors to learn how to prevent them.

Bruce Budowle, a former F.B.I. scientist now at the University of North Texas’s Health Science Center, said, “For the FBI to say, ‘It’s the states’ problem’ and that’s the end of it, that’s the ostrich burying its head in the sand and not contributing to improvement of the process.”

As to the 166 criminal cases, Mr. Scheck, who is part of New York’s commission for forensic science, said it was unclear how many had been considered closed by law enforcement and ought to be reopened to see if the DNA pointed to other suspects. He said the errors underscored the need for more information-sharing among entities in the criminal justice system to ensure that new DNA evidence received consideration even if it appeared years after someone had already been convicted of a crime.

One of the errors involved the analysis of DNA found on a gun that had been handled by several people and used in a shooting in the Bronx in 2010. A man, Marc Outram, was indicted on the basis of eyewitness testimony from a police officer, even though Mr. Outram’s DNA profile was deemed as conflicting with the DNA mixture found on the gun, the authorities said. Following the discovery of an analyst’s error, the authorities again examined the DNA from the gun, this time discerning a different DNA profile. The results again excluded Mr. Outram, and the new DNA profile appeared to match another person in the database.

Bronx prosecutors plan to drop the case against Mr. Outram, said Steven Reed, a spokesman for the Bronx district attorney’s office, adding that the case was “going to be dismissed for reasons other than the problem with the DNA profile.”

A version of this article appears in print on January 25, 2014, on page A15 of the New York edition with the headline: F.B.I. Audit of Database That Indexes DNA Finds Errors in Profiles.