Post-mortem interval, PMI, or the time of death—whatever you call it, it’s a critical forensic factor in death investigations. The growth of insects or bacteria, the chemistry and color of the blood that has ceased to flow, even the composition of the smell of death itself have been avenues taken toward pinpointing that elusive last heartbeat.

The changes wrought in RNA of certain tissues in the recently-dead can be a critical marker to determine the PMI, according to the latest entry, a paper by a Spanish and Portuguese team in the journal Nature Communications.

By comparing the RNA transcriptome both before and after death, and in dozens of different body tissues, they have identified markers that show predictable and consistent changes at the level of molecules.

“Our analyses suggest that the patterns of gene expression change with time after death in a tissue specific manner, and might thus be collectively used to predict the PMI for a given individual,” they write. “By comparing ante- and post-mortem blood samples, we identify the cascade of transcriptional events triggered by death of the organism.”

The samples were from the Genotype-Tissue Expression Project (GTEx), a project launched by the National Institutes of Health in 2010.

The data included 7,105 samples from 540 donors, which were taken from 36 tissue types.

The GTEx data allowed them to plot the chemical changes in each of the tissues for each of the donors.

Finally, they split the data for machine-learning purposes. Seventy-five percent went into training the computer, and the remaining 25 percent of the samples were used to test its predictive accuracy.

They found it was able to narrow down the PMI to statistically-significant ranges. 

“Based on tissue-specific response of the transcriptome to PMI, we built machine-learning models to predict the time of death of a recently-deceased individual,” they write. “We show that RNA-seq performed on a few key tissues could become a powerful tool to aid in forensic pathology.”

One of the key changes across 13 tissues was RNASE2, from the ribonuclease family, which shows a consistent decease in expression. Other markers included two alpha globin genes (HBA1 and HBA2), which are involved in transport of oxygen from the lungs to other tissues, and also several histone genes, which showed increased expression in death.

The forensic biological clues could even extend beyond the PMI, they add. 

“It could carry the footprint not only of the time since death, but also of the cause of death—even though we could not properly carry out these analyses because of the small sample sizes available,” they write.

Pedro G. Ferreira, the lead author, alumnus of the Center for Genomic Regulation in Barcelona and currently at the University of Porto in Portugal, said the results show forensic promise.

“We found that many genes change expression over relatively short post-mortem intervals, in a largely tissue-specific manner,” said Ferreira, in a statement. “This information helps us to better understand variation and also it allows us to identify the transcriptional events triggered by death in an organism.”

Genetic changes have been the focus of previous PMI studies. For instance, a 2015 paper in the journal Frontiers in Genetics looked at the DNA methylation changes in pig skulls kept in tightly-controlled laboratory environments. Another in a 2013 issue of Forensic Science International probed into tooth pulp to find RNA degradation, and corresponding morphological changes.