Trying to determine the time of death for a murder victim has never been an exact science, but that could all be about to change.
While forensic science has managed to drastically improve the accuracy of determining the time of death when compared with just a few decades ago, being able to tell when someone died to the hour would be an incredible leap forward for the fields of science and criminology.
So, news that the latest genetic science could help bring us one step closer to this reality is surely cause for cautious optimism.
According to Science, the breakthrough was made by computational biologist Roderic Guigó from the Centre for Genomic Regulation in Barcelona, who is part of a consortium of geneticists and molecular biologists trying to understand how the body makes different cells perform different functions.
To this end, Guigó and his colleagues collected 9,000 samples of human tissue, each of which included data telling the researchers when the person died and when the sample was collected and preserved.
After analysing these samples, it could be seen that every one of them differed in the number of distinct patterns of gene activity, which could potentially be used to look back and find a precise time of death.
In fact, how the body responds to death is quite tissue-specific, with the samples showing that while more than 600 muscle genes showed a quick increase or decrease in activity after death, the same was not seen in brain tissue, for example.
Power of AI
At this point, the team introduced the power of machine learning by using specially developed software to learn these genetic activity patterns of just under 400 people and see if it could pinpoint the time of death of the people behind the samples.
After it had churned through the data, the software found that in one example, there was a noticeable decrease in the number of genes involved in DNA production and immune response, but an increase in those responsible for stress.
This, the software calculated, put the person’s time of death to six hours before preservation, with the majority of gene activity occurring between seven and 14 hours after death.
Challenges for the team going forward include the fact that it has not been able to reduce the number of genes needed to make an estimate, which would make the process too expensive to run on a regular basis in its current state.
While admitting that it is so far just an “academic exercise”, Guigó is optimistic about its chances to become more feasible.