DNA, however, is not without its challenges. Poor-quality samples or partial profiles containing degraded DNA, for example, can make it difficult to identify the DNA donor. So too can complex DNA mixtures containing genetic data from multiple persons.
The recent introduction of sophisticated forensic DNA interpretation software greatly enhances both the power and reliability of DNA profiling. The software packages (one of which I and two other civil servants developed) enable forensic analysts to contribute to criminal investigations in which DNA evidence previously was considered too complex to interpret. They also vastly improve the ability to analyze low-level, degraded or mixed DNA samples with a significantly higher degree of accuracy than was previously possible.
While these software packages differ from each other in detail, all were developed to use information in the DNA profile that was discarded in the past. Doing so improves interpretation, particularly with respect to complex DNA mixtures previously thought to be unresolvable.
Forensic DNA interpretation software already has been used in thousands of cases worldwide. A recent Michigan case, for example, used such a tool to provide key evidence in an armed-robbery case, interpreting the profile obtained from sweat inside a sneaker left at the crime scene. Forensic software's ability to separate multiple profiles of DNA extracted from evidence has also helped to crack cold cases, such as a 2009 murder recently solved by the Sarasota County, Fla., Sheriff's Office.
But while increased acceptance and use of forensic DNA software is likely to have a major impact on prosecution and case-clearance rates, as well as potentially serving as a deterrent to crime, it is not without its detractors. Some critics assert, for example, that DNA software is too new to rely confidently on its results. While it is true that sophisticated DNA software has been in use for less than a decade, the probability models and other statistical-analysis methods that it employs actually date to World War II and are widely used in everything from computational biology and weather prediction to physics, engineering and the stock market.
Similarly, questions have been raised about whether DNA software has been properly reviewed and validated. Peer-reviewed scientific journals, in fact, have published numerous scientific papers about sophisticated DNA software. Internal validations also have been carried out by all labs in current casework (as required by their accreditation), while both the International Society for Forensic Genetics and the Scientific Working Group on DNA Analysis Methods have published validation guidelines.
It is important to note, however, that forensic professionals require access to training and professional support. As a result, there is at least the possibility for cases which depend on forensic DNA software to be undermined by poor training or poor support.
Also of concern is the fact that algorithms used in DNA software are sometimes secret, leaving attorneys to deal with a "black box" situation -- that is, data goes in, a solution comes out, but there is no explanation of what happens in between. This, however, doesn't apply across the board. Developers of most forensic software now divulge their code under non-disclosure agreements. In addition, some of the software is open-source, making codes and algorithms readily available for users to explore and improve.
Clearly, potential issues must constantly be considered and evaluated. I would contend, however, that the benefits of forensic DNA software far outweigh any drawbacks. Recognizing that sophisticated forensic software, like any tool, has its limitations ultimately will enable DNA evidence to evolve into an even more effective tool for law enforcement and criminal justice.
