RENEE MONTAGNE, host:
Law enforcement in this country often looks to science for help in solving crimes. But using cutting-edge science as evidence can be problematic. A study out this week highlights that. It describes the genetics the FBI used in its investigation of the anthrax letter attacks back in 2001.
NPR's Joe Palca has more.
JOE PALCA: The FBI says its investigation of the anthrax letters was one of the largest ever undertaken. In 2008, the Justice Department was about to indict a researcher named Bruce Ivins for committing the crime, but Ivins took his own life.
A key piece of evidence tying Ivins to the anthrax attacks was the ability to match the anthrax bacteria in his lab to the bacterial spores found in the letters. To understand how the FBI made that connection, you have to know something about how the anthrax bacteria grow in the lab.
Dr. PAUL KEIM (Northern Arizona University): They go from one cell to nearly a billion cells overnight.
PALCA: Paul Keim is a geneticist at Northern Arizona University and T-GEN, a non-profit biomedical research foundation. Those billion cells are what microbiologists call a colony.
Dr. KEIM: It's a group of bacterial cells that have grown up from a single cell.
PALCA: If every batch of bacteria produced identical colonies, there would be no way to tell them apart. But it turns out each batch produced some colonies that looked slightly different from one another. So Keim says scientists working with the FBI looked at the genomes of these different colonies.
Dr. KEIM: And what they found was is there were subtle differences in their genome that affected they way they grew on a plate.
PALCA: Those differences gave a kind of genetic fingerprint for each batch of anthrax. Now, the FBI had a repository of a thousand batches of anthrax bacteria from different labs. The new fingerprinting technique, reported in the latest issue of the journal PNAS, let the FBI ask a crucial question.
Dr. KEIM: Is the DNA fingerprint that they saw in the letter spores observed in any of these repository samples? And they found them.
PALCA: In fact, all the bacterial samples in the letters matched the fingerprints of samples taken from Bruce Ivins' lab. That sounds pretty conclusive. But here's the problem: Paul Keim says scientists don't yet know if the letter samples might match other batches, as well. It's not certain each batch has a distinctive signature.
Dr. KEIM: We're going to go and look. We're going to go look and see if there is, in fact, a unique genetic signature in every single batch. Theory says that there should be.
PALCA: Testing theories is what science is all about. But how rock-solid does a theory have to be before you can use it in the courtroom? Stanford Medical School's David Relman was vice-chair of the National Academy of Sciences panel that looked at the science the FBI used in its investigation.
Dr. DAVID RELMAN (Stanford Medical School): There's always a tension between wanting to embrace and integrate a new approach quickly, and yet wanting to be sure that it's done properly and properly validated.
PALCA: Another problem the Academy panel wrestled with is how a jury responds to scientific evidence. Alice Gast is president of Lehigh University. She chaired the panel.
Dr. ALICE GAST (Lehigh University): One of the things that concerns me when science meets law enforcement is that science is sometimes put up on a pedestal that's a bit higher than other pieces of evidence.
PALCA: She worries calling something scientific makes it sound more definitive than it really is. The FBI says science was only one part of its case against Ivins, but it clearly was an important part. And weighing scientific evidence is tricky for scientists, let alone members of a jury.
Joe Palca, NPR News, Washington.