Saturday, June 29, 2013

Criminal DNA Testing

DNA testing under the microscope

With the proliferation of CSI type crime dramas over the last few years, many people from Juries to the lay public have come to assume that DNA testing is a quick and foolproof way of proving the guilt or innocence of a criminal suspect. In fact, some experts even worry that this “CSI culture” is invading the purview of the courts; leading juries to expect and even demand DNA evidence from the prosecution even in cases where such evidence would add little of substance to the analysis. Adding to the confusion are the range of media reports, some of which warn about the possible inaccuracies of DNA testing procedures and others which hail the ways in which the technology has been used to absolve and free convicted criminals.

However, despite the media hype and many public misconceptions to the contrary, DNA testing is a highly reliable and accurate method of matching a sample with a donor, or of proving that no match exists. But, this accuracy depends entirely on the methods used, the care and skill of investigators and lab specialists, and the presentation of the resulting evidence by lawyers during trial.

What is DNA testing?

DNA testing is the process of comparing a sample of DNA, such as from a crime scene, to a potential suspect; a match implying the suspect’s presence at the scene. This is possible because every person’s DNA is unique, distinct in some way from every other person who has ever lived or who will live – with the exception of identical twins who share exact DNA. However, while each individual’s DNA is unique to that person, current testing methods do not allow for a complete DNA profile.

Methods are improving however. First developed in 1985 by British scientist Alec Jefferys and colleagues, DNA testing has come a long way from the early days, both in quality and cost. Analyses that once required a fresh tissue sample the size of a quarter can now be completed with just a few partially degraded cells from a cold case file. Tests which were once impossible, such as distinguishing individual DNA profiles from the mixed semen of multiple rapists, have now become common place and DNA testing has even been used to compare non-human samples, such as the seeds of a tree which were used to tie a suspect to a crime scene in one recent Arizona case.

How accurate is a match?

The entirety of a DNA sample is not currently tested. Instead, sections known to vary substantially among humans are sampled and compared. While any single sample site has about a 7.5% chance of matching another random human, insufficient for criminal identification, the odds increase exponentially with every additional sample site considered. At about 4 to 6 sample sites – called markers – identification becomes reasonably possible. For good measure, the standard FBI test compares 13 markers, leading to greater than 1 in 1 billion certainty that a suspect is in fact the person who left the sample. In other words, by comparing 13 quality (which means highly variable) markers, less than 7 people in the entire population will be statistically probable matches. Given that the pool of suspects cannot ever include the entire population of the planet, this level of certainty can provide the evidence necessary for conviction beyond a reasonable doubt.

(For more details on these numbers please visit here)

What’s the catch?

Naturally, there are some caveats about the numbers listed above. While the statistics are solid, they have been tested repeatedly by many different mathematicians and scientists the world over, the accuracy of any purported match comes down to process; and here is where CSI type shows drive unreasonable expectations about DNA results. As with much of the criminal justice system, human error is a major factor. Collecting DNA samples from a crime scene often requires expert training and professional equipment. Samples must then be transferred and stored in approved ways to avoid degradation. A qualified lab must properly utilize expensive edge equipment to analyze the sample to produce a DNA profile. That profile is then compared against either a sample taken from a suspect or against one of the many available databases of DNA profiles, which imply a number of privacy and security concerns, and a match determined.

Every step of this process is fraught with potential error. Crime scenes can be contaminated or evidence planted, samples can become contaminated or degrade, labs might employ outdated techniques or scientists might not have the requisite training, computer error could be introduced, jurors might give too much weight to DNA evidence; and all along the way, human malfeasance might undermine the validity of a match. And all this stems from but a single situation; a single trial.


While for individuals the results of a single DNA match are of the utmost importance, whether the evidence is being used to convict or exonerate, for the criminal justice system – and its promise of fairness and justice for the public as a whole – systemic problems become the more critical consideration. On this level, a number of issues must be dealt with if DNA testing is to remain a viable tool in criminal prosecutions.

Evidence from analysis of the UKs national DNA database shows that a disproportionate number of the DNA profiles in the system come from minorities. While it can be argued that this represents the increased prevalence of minority convictions, the increasing use of so called “cold matches” wherein DNA from every arrested person is periodically checked against the database, leads to possibilities of institutionalized racism and prejudice. Additionally, the increasingly popular practice of collecting samples from arrestees prior to any conviction is leading to a number of very serious privacy and security considerations. DNA samples can, in addition to providing material for a possible match to a criminal investigation, also reveal family connections, expose health conditions, and even predict sexual orientation or inclinations.

Protecting the future

DNA profiling has become an increasingly important feature of our criminal justice system, both exonerating the innocent and convicting previously unreachable suspects; however its future use is imperiled by a growing number of developing problems with the practice. More must be done to protect this valuable criminal justice resource from corruption, error, and prejudice.  The worrisome backlog of untested samples should be cut down, and the public at large – many of whom ultimately become jurors – must be better informed about the accuracy, and potential shortfalls, of the science.