Department of Justice and Defense grants to be used to improve forensic identification

The U.S. Department of Justice along with the Department of Defense has awarded Catherine Grgicak, PhD, assistant professor in biomedical forensic sciences at Boston University School of Medicine, approximately $2.5 million to more accurately analyze DNA evidence at a crime scene.

Grgicak, along with colleagues from Rutgers University-Camden and the Massachusetts Institute of Technology, are working to develop algorithmic methods and models to aid in complex DNA mixture interpretation, including identifying the number of possible contributors in an evidence sample and determining the strength of a DNA match.

DNA evidence often is simplified in modern culture on television shows such as CSI, but the forensic science community has been embroiled in a debate for many years about how to analyse samples that may be obscured by DNA from people other than the subject. Complex mixture profiles are those that contain DNA originating from more than two contributors, or from samples where only a few cells were collected from the scene.

Grgicak and her collaborators are producing and testing methods that provide match statistics for DNA comparisons as well as determining the probability that a certain number of contributors give rise to biological evidence. "Through statistical analysis of data, we are hoping to examine the degree of accuracy and reliability of methods used by forensic scientists in order to achieve a more complete understanding of the basis for interpretation," explained Grgicak. The continued development of these solutions, she believes, is expected to benefit the forensic DNA discipline and will lead to the development of new methods or protocols to interpret forensic data from physical evidence.

Grgicak and her colleagues already have developed a computational tool to identify the number of contributors to a forensic DNA sample called NOCIt. According to Grgicak, this enhanced approach to DNA mixture interpretation applies statistical and computational techniques that can accurately assess complex, noisy evidence samples involving many contributors and could gradually become incorporated into standard practice in crime labs all around the world.

Provided by Boston University Medical Center