Improving mixed-stain DNA analyses
With routine forensic DNA research methods, you can produce reliable DNA profiles from as few
as 100 individual cells. However, there is an increasing demand for even more sensitive methods as, increasingly, the crime scene samples presented for DNA analysis are minute. This introduces two major challenges:
- It decreases the evidential relevance of each individual sample because the smaller the sample
– in terms of the number of cells – the higher the probability that its presence at the crime scene is irrelevant, either due to pre-crime or post-crime contamination. - Increased routine sensitivity also leads to the identification of mixed-stains (i.e. biological crime scene stains that contain DNA from multiple individuals.
At the moment, we use multiplex PCR kits that are available commercially to identify routine DNA profiles. However, these kits have one major disadvantage. When biological crime-scene stains contain DNA from multiple donors (e.g. a cigarette smoked or a bottle used by several people), the minimum DNA contribution in this mixture can only be detected if it amounts to about 5% or more of the total amount of DNA. Because of this restriction, technically called preferential amplification, it is impossible to analyse in detail many of the crucial mixed DNA stains and, in many cases, crucial DNA evidence is lost.
We estimate that the loss of this sort of information leads to the exoneration of suspects in at least
10-20 cases per year. While this is not much percentagewise, the public perception of such “lost cases” can be very damaging. We plan to substantially improve the detailed analysis of biological crime scene stains from multiple donors with highly-sensitive and high-throughput next-generation whole genome sequencing technologies. This technology has already been proven in cancer diagnostics. We plan to take this technology to the next level and develop methods that will enable us to routinely identify individual components in mixed biological crime-scene samples irrespective of knowledge about the number of contributors.
