24 December 2015

Fourth Sunday of Advent: Number crunchers as crime fighters

Forensic genetics

Getting away with “the perfect crime” has become ever more difficult thanks to new DNA analysis techniques. Behind these techniques is a whole lot of IT, statistics and mathematics, all of which help to make calculations as accurate as possible. Today’s tale is that of two researchers who play a part in capturing criminals.

Steffen Lauritzen, Professor of Statistics, Department of Mathematics.
Field of Research: Mathematical Statistics                                           

Therese Graversen, Postdoc in Statistics, Department of Mathematics.
Field of Research: Statistical methods and software development in forensic genetics.

The human genome was first sequenced in 2001, at a cost of billions of kroner and after many years of work to decipher the DNA profile of human beings. Since then, the technique has developed considerably and become incredibly cheaper, much faster and more accurate. The use of DNA analysis has been used in fields as varied as prehistoric human migration and criminal justice. DNA techniques used in criminal cases means that successfully committing “the perfect crime” is becoming ever more difficult. For example, police and investigators no longer need to scan a crime scene with flashlights to look for fingerprints. The authorities can now use modern techniques where the tiniest strands of hair or minute amounts of bodily fluids can reveal a culprit, and be used as technical evidence in a court of law. It used to be that formed by hypothesis and heavy suspicion were used in trials. Finally, the accuracy of the DNA method has also been decisive factor in freeing wrongly accused death row inmates.

Math and crime

Biologists aren’t the only ones involved with the development and use of criminal justice related DNA techniques. Without IT and math, the techniques would never have evolved to become cheap, fast and accurate. Subsequently, this Christmas tale doesn’t revolve around detectives, DNA researchers or biologists, but a pair of statisticians. Both of them assist the police and other government officials to solve crimes.

Both researchers in today’s Christmas crime fighting tale, Professor Steffen Lauritsen and Postdoc Therese Graversen, work at the University of Copenhagen’s Department of Mathematics.

Therese Graversen:

“I conduct statistics research. Among other things, I develop statistical methods and software for forensic genetics.

Statistics becomes an important factor when DNA evidence is introduced to a criminal case to help identify a person. In a straightforward case, statistical analysis serves to calculate the probability of finding another person with the same DNA profile. However, when dealing with poor quality evidence or evidence that has the DNA of numerous people, the statistical analysis soon becomes quite complex. One of the main problems is to keep the necessary computer calculations down to a manageable level so that they can actually be conducted in practice,” says Therese Graversen about her work.

Steffen Lauritzen:

“My research is typically applied to a number of areas tangential my field. But I am most known for my work with graphic models, a statistics sub-field that I have in large part established and developed. Graphic models are statistical models that describe the relationship between variously sized things by representing them with a mathematical graph – or a network, if you like – that describes how things are directly or indirectly related to one another. There are various graph types, each with their own interpretation.

First and foremost, my research is used to help understand complex relationships in the world around us. For example, graphic models can be used for telecommunications, for various types of advanced systems that might play a part in the surveillance of technical processes, for searching and interpreting information from the internet, for the interpretation of trace DNA in criminal cases, the identification of genetic networks and much more. The most exciting thing about being a statistician is the variety of topics and fields that one comes into contact with through my own field’s applications, and in part by being a heavy user of a great many types of advanced mathematics,” underscores Steffen Lauritzen about his work.

So what about evidence in criminal cases and statistics?

Therese Graversen:

“When one looks at evidence, whether in the lab or during the judicial process, it is always natural to place a question mark over the extent to which statistical methods are suitable for the case at hand. It is precisely these types of questions that I want to contribute an answer to. But I like the idea of solving a few problems that other professionals have to deal with on a daily basis. It is also fascinating to get the chance to challenge the way people think about given situations.”