It's a trope of prime-time detective shows: a man is about to be convicted of a brutal murder when the detectives find the weapon, complete with traces of DNA. They send the DNA for analysis and in just a few TV minutes, they have their answer. The man is innocent. Case closed. Unfortunately, this level of scientific certainty is just as fictional as the show's plot. DNA is hardly the airtight evidence most people think it is.

It All Started So Well

In a field plagued by junk science, DNA analysis has always been held up as unassailable. It was conceived in 1984 by British geneticist Alec Jeffreys, who accidentally discovered it while researching genetic sequencing. He soon used it to help nearby police crack two unsolved murder cases, which made his creation front-page news around the world. Soon, DNA typing was an important part of the criminal justice system, helping both overturn wrongful convictions and prove guilt beyond a reasonable doubt. But at the same time the technique was gaining in popularity, it lost a good amount of its precision.

Here's why: Alec Jeffreys' technique only worked when you compared one large sample of DNA to another large sample of DNA — say, to determine whether a pool of blood found at the crime scene came from a particular suspect. As the science progressed, labs gained the ability to use smaller samples, such as those from fingerprints or specks of saliva, and mixtures that included DNA from multiple people. That was handy for law enforcement, but it made the job of scientists much, much tougher.

An analysis of DNA.

10 Scientists, 10 Different Answers

You may recall that all humans have 99.9 percent of their genes in common. It's that last 0.1 percent that geneticists look for in DNA analysis. Specifically, there are certain pairs of genes, or alleles, that vary from person to person, and the most accurate way to identify whether DNA came from a certain person is to compare samples at as many locations as possible. The Atlantic explains, "When a forensic analyst compares alleles at 13 locations — the standard for most labs — the odds of two unrelated people matching at all of them are less than one in 1 billion."

But when there are tiny samples, degraded DNA, or mixtures, it's not that simple. You have to decide how many people's DNA are involved, which alleles belong to which person, and identify whether alleles have disappeared. Suddenly, interpretation comes into play. That's where things can go very wrong.

For one study, 17 lab technicians were asked to re-analyze a DNA sample which, unbeknownst to them, came from a rape trial where the defendant was found guilty. Only one of the 17 decided that the defendant was guilty — the rest concluded that the DNA was either inconclusive or didn't belong to the defendant.

As forensic scientist Peter Gill famously said in 2005, "If you show 10 colleagues a mixture, you will probably end up with 10 different answers." This isn't to say that DNA is useless — in trained, competent hands, DNA analysis can be very useful. But many juries and courtrooms see it as scientific perfection, and it is anything but.

Why We Can't Always Trust DNA Evidence

Written by Ashley Hamer February 16, 2017

Curiosity uses cookies to improve site performance, for analytics and for advertising. By continuing to use our site, you accept our use of cookies, our Privacy Policy and Terms of Use.