Quantitative color analysis of burned bone to predict DNA quantity, quality, and genotyping success

Badly burned skeletal remains are commonly submitted to forensic laboratories for victim identification via DNA analysis methods. Burned skeletal remains present many challenges for DNA analysis as they can contain low amounts of DNA which can also be damaged and degraded, resulting in partial or no STR profiles. Therefore, a simple, but effective screening method that identifies which samples may provide the most successful STR or mtDNA typing results for identification would enable forensic laboratories to save time, money, and resources. One metric that can be used and a screening method is the color of burned bone, as bone color changes with exposure to fire as temperature and length of exposure increase. This research developed a quantitative screening method based on the surface color of burned bone. The different visual bone colors (light brown, dark brown, black, gray, and white) were quantified using the Commission on Illumination L*a*b color space. These values were then compared to DNA yield, STR, and mtDNA profile completeness to identify whether the L*a*b values can predict genotyping success. A Bayesian network was constructed to determine the probability of STR typing success, given a set of L*a*b values. Results demonstrated that samples with an a* value greater than or equal to one and b* value greater than eight (light brown and dark brown burned samples) were the most predictive of STR typing success and mtDNA typing success. A decision tree for processing burned bones was constructed based on the color value thresholds.