DNA evidence has done extraordinary things for justice. It has freed hundreds of wrongly convicted people, identified remains, and solved cold cases that everyone had given up on. The Innocence Project’s exoneration record alone is enough to establish that DNA testing, used carefully, is among the most powerful tools forensic science has. That’s the part that’s true. The part that’s overstated is the cultural assumption that a DNA match settles a case. Modern DNA technology is sensitive enough that the question is no longer “is this person’s DNA present?” but “what does its presence actually mean?” — and that question is much harder.
The science is still solid. The interpretation is where things go wrong.
Touch DNA and the transfer problem
Modern testing can amplify DNA from just a few cells, which is impressive but introduces the transfer problem. Your DNA can end up at a crime scene without you ever being there. It can transfer through a handshake, a shared tool, a doorknob, a piece of clothing, a paramedic’s gloves. Studies have shown DNA from one person ending up on objects another person never touched, sometimes hours or even a day later. In one Australian case, a man’s DNA was found at a murder scene he could not have been at — paramedics had attended both events using equipment that wasn’t fully cleaned. As testing sensitivity has gone up, the meaning of “your DNA was there” has gone down. Presence is not necessarily contact, and contact is not necessarily relevance to the crime.
Mixtures and the statistics problem
When DNA from multiple people is on the same sample — a doorknob in a high-traffic place, a weapon handled by several people — interpretation gets dramatically harder. Probabilistic genotyping software like STRmix and TrueAllele can sometimes deconvolute mixtures, but their statistical conclusions vary depending on assumptions made by the analyst. Different labs can reach different match probabilities on the same sample. In a 2018 study, multiple labs analyzing the same mixture produced widely divergent likelihood ratios. Juries hearing “1 in 1 quadrillion” don’t necessarily understand that the underlying number depends on choices that another competent analyst might have made differently. The certainty of the headline number outruns the certainty of the underlying analysis.
Database, contamination, and chain of custody
Cold-hit DNA matches from law-enforcement databases sound conclusive but raise their own issues. The “birthday-paradox” math means very large databases can produce coincidental matches even with rare profiles, especially partial ones. Lab contamination has caused real problems — Germany famously chased the “Phantom of Heilbronn” for years before realizing the same DNA appearing at dozens of crime scenes belonged to a worker at the swab factory. Chain-of-custody errors and sample swaps still occur. Each of these failures is rare, but at the scale of the modern criminal-justice system, rare adds up to many real cases.
The bottom line
DNA evidence remains powerful, especially for exclusion. As inclusion evidence, treat it as one important piece of the puzzle, not the puzzle itself. Match probabilities are conclusions about populations, not certainties about individuals.
Leave a Reply