Forensic Fails: Why DNA evidence doesn’t always provide the correct answers

Your DNA says way more about you than a finger print, cigarette stub, boot print outside the window or any of the other evidence Sherlock Holmes always found so informative. DNA testing has been a great boon for criminal investigators, but is it always compelling evidence? Here are some notable occasions where the silver bullet that is DNA evidence has got jammed in the barrel of the gun of justice and messed everything up:

“Oh no, my justice gun!”

Evil twins

Evil twins are the most obvious example of a pitfall of DNA evidence. Twins which originated from the same zygote (fused egg and sperm cells) have identical DNA and are thus identical twins. Fraternal twins are from separate eggs which are fertilized simultaneously by different sperm. In the case of Ronald and Donald Smith, DNA evidence at the scene of a murder suggested that one of the two identical twins was responsible, but not which of them it was. The fact that Donald had an alibi didn’t help – the two of them looked too similar. In this instance good old fashioned finger print evidence – one thing that even identical twins don’t share – determined that Ronald, not Donald was guilty.


A chimera is a person who doesn’t just have one genome: they have two or maybe more. This might sound bizarre, but it’s a more common occurrence than you might think:

Transplant recipients are chimeras: receiving another person’s organ means you receive their cells, carrying their genome. These cells will continue to divide and replicate by themselves and will never adopt the host’s DNA. Although no cases have been recorded to date, this means that theoretically a bone marrow transplant recipient could leave blood at a crime scene but not be linked to it if they then provided the police with a saliva swab for testing.

Natural chimerism was once thought to be rare but now appears to be quite a common phenomenon: most women who have been pregnant swap small numbers of cells with their child. These cells stick around and, even years later, can be found all over the mothers body – even in her brain. Chimerism can also occur just after conception, where two fertilised eggs (that would otherwise have become fraternal twins) fuse together to create one embryo, containing cells with two different genotypes.

The most bizarre legal problem with chimerism to date is probably the case of Lydia Fairchild. Lydia was pregnant with her third child and applying for financial support. This required her children had a paternity test to confirm that her ex-partner was their father. When the results came back, they concurred that he was the children’s father…but she was not their mother. Adamant that she had given birth to the two existing kids yet facing kidnapping charges, Lydia demanded an official be in the room with her when she gave birth to child #3. After delivery a maternity test was immediately performed and once again showed that the child she had just given birth to was not hers! After much confusion and specialist testing it was discovered that Lydia had once had a fraternal twin sister…who she had engulfed in the womb. All that remained of the twin was her ovaries and womb which were inside Lydia, making their own eggs. Lydia may have given birth to these children, but she was not their mother; she was actually their aunt.

The kidnapping charges were dropped.

Human Error

Just as any crowd is only as clever as its stupidest member; any forensic test is only as reliable as the technician performing it. The number of incidents of lost, mixed up or contaminated DNA evidence leading to wrongful convictions or acquittals is shocking, but thankfully appears to be declining as testing standards improve.

A particularly stupid case of human error was the Phantom of Heilbron, or ‘The Woman Without a Face’. This woman was never seen by witnesses or caught on cameras, and yet over 15 years her DNA was found at the scenes of 40 violent crimes, including 6 murders which took place all across Germany. Although a €300,000 reward was offered for evidence leading to the identification of this possible serial killer, German police cautioned that she was far too dangerous to approach. After many years, investigators concluded that the Phantom’s DNA actually belonged to an employee of the factory that made the cotton swabs they used to collect samples at crime scenes. A nearby precinct which used swabs from a different supplier had not picked up any of the Phantom’s crimes, despite being in the heart of her territory. Most of the murders were later resolved.

What will we do wrong next?

Where can forensic genetics go from here? One research group has recently published their preliminary findings in a project trying to predict people’s facial structure based on their DNA (Link to paper, Link to open-access article). The study used high resolution mapping of the faces of 600 individuals to identify genes significantly involved in determining face shape. 24 small mutations in 20 different genes were identified as having predictive use, but as you can imagine, drawing a face based on only 24 suggestions of cheekbone height, eye shape, etc can’t really match anything done by a sketch artist (although it becomes more useful when you combine it with data from genes for eye and hair colour, or ancestry markers). I find the idea inherently creepy, so I’m glad we’re quite far off from making this a reality.

It seems that forensic genetics is yet another example of something that is not totally understood being too readily adopted and relied upon. Although we are beginning to spot the pitfalls of DNA evidence and look at it critically, this should clearly not be seen as a watertight tool. Considering all the mishaps we have seen with something that should be as relatively straight-forward as DNA, the recent introduction of fMRI for lie-detection in the United States law enforcement makes me incredibly uneasy; this is a poorly understood technology being used to make assumptions about a poorly understood organ. Mounting evidence suggests that the activity measured by fMRI doesn’t correlate with which parts of the brain are actually being used, and furthermore, location and function are only loosely related in the brain – particularly for complex tasks, such as creating a lie*. Even IF fMRI could reliably tell that I was using the ‘creative centres’ in my brain while I recounted my alibi, what does that even mean? Who is to say I wasn’t just looking for the right adjective? This guy really brought the incredulity of this technology home when he used it to demonstrate that a dead salmon he brought in a shop could empathise with photos of humans in different social situations.

Technology is fallible, science is fallible and people are fallible. There is no such thing as conclusive evidence of anything, fMRI sucks.

*I’m sorry I give no peer reviewed sources for my fMRI rant – please refer to this better informed blog, which does contain appropriate links

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s