Study On Inferring The Postmortem Interval Of Boiled And Buried Rat Carcasses By Microbial Community Succession

Objective:The estimation of postmortem interval(PMI)has always been regarded as a hot and difficult issue in forensic research.There have been studies using microbial community succession rules to infer the postmortem interval of corpses exposed to air and water.But in some cases,murderers use boiled or buried methods to process corpses in order to conceal criminal evidence.Currently,there is no reliable PMI inference method for boiled corpses.Buried corpses are often highly corrupt or skeleton when found due to strong concealment characteristics.In this study,rat models were prepared by Simulating corpse boiled and buried cases,and rectal and bone samples were collected respectively to detect the relationship between microbial community and postmortem interval,with a view to providing valuable references for inferring the postmortem interval in these two types of cases.Methods:Preparation of models:Simulated boiled corpse cases,there were 8 rats in the experimental group were sacrificed and boiled in water,while 8 rats in the control group were sacrificed only for cervical dislocation.Rectal samples were collected at 9 different time points in an artificial climate box.Simulated buried cases,there were 20 rats were buried in a homogeneous soil pit with a depth of 20 cm.Samples of the upper limb humerus and lower limb femur were collected at 5 different time points.Extracting DNA,amplifying bacterial 16S r RNA genes by using universal primers,constructing a library,purifying with magnetic beads,quantifing the library,and sequencing on the Illumina Miseq platform.Using software such as QIIME and USEARCH to merge,quality filter,and cluster sequencing data,comparing them with the Silva database for species annotation.Conducting statistical analysis based on the R software platform to study the changes in bacterial community composition and abundance,α-Diversity analysis,β-Diversity analysis,and Principal coordinates analysis(PCo A)during the decomposition process of corpses.Using machine learning algorithms to perform regression analysis on microbial communities and PMI to build a random forest model for predicting PMI.Results:There were significant differences in the composition and relative abundance of bacteria in the rectum of rats boiled and unboiled in an artificial climate box during the decomposition process of corpses,but theirα-Diversity have shown a downward trend,and the similarity of microbial communities have a significant linear relationship with postmortem interval.The goodness of fit of the two groups of rats constructing random forest models based on microbial markers reached 68.00%and 84.00%,respectively.The mean absolute error(MAE)within 45 days of decomposition was 2.05±0.41d and 1.48±0.31d,respectively.During the 60 days decomposition process of buried rats,skeletal bacteria exhibited different performance at different time points from the phylum and class levels,and there was a significant linear relationship between microbial community similarity and postmortem interval.The R~2of random forest model for predicting the time of death based on all OTUs levels and the selected microbial marker set reached 85.07%and 86.32%,respectively.The MAE was 2.20±1.80 d and 2.11±1.81d,respectively.Conclusion:This experiment is based on rat models to simulate boiled corpses and buried corpses.Although they undergo different disposal methods(whether they are boiled or unboiled)and different environments(indoor environment or soil buried),they all exhibit the same significant linear trend between bacterial community succession and PMI.Using random forests to establish postmortem time inference models has good predictability,which provides references for estimating postmortem time based on the succession rules of microbial communities in cases of boiled corpses and buried corpses.