"How reliable is DNA in identifying suspects?"

That's the title of this very interesting article from the Los Angeles Times. The answer: maybe not anywhere near as reliable as the FBI wants you to believe. I don't fully understand the science and statistics discussed in the article, but the problem appears to be that the FBI's DNA match methodology is based on some assumptions that have never been subjected to empirical testing. More troublingly, the article talks about the FBI's efforts to block any further inquiry into the issue.

Link via How Appealing.