基于埋葬大鼠的骨骼微生物群落演替规律推断死亡时间

Estimation of postmortem interval based on the succession of skeletal microbial communities in buried rats

目的 基于大鼠模型研究埋葬尸体骨骼微生物的群落演替规律,并以此推断死亡时间(PMI),为埋尸案件的死亡时间推断提供参考。方法 将20只大鼠埋葬于深度20 cm的土壤中,分别于5个时间点(7 d、14 d、30 d、45 d、60 d)采集上肢肱骨及下肢股骨,对细菌16S rRNA基因测序及数据分析,利用机器学习算法构建预测死亡时间的随机森林模型。结果 埋葬大鼠在60 d的分解过程中,上肢肱骨和下肢股骨微生物群落无显著差异,不同时间点微生物群落差异显著。在尸体腐败进程中,骨骼厚壁菌门、拟杆菌门相对丰度总体呈现上升趋势,变形菌门、放线菌门呈现下降趋势。利用随机森林算法筛选出与PMI相关的89个OTU作为标记物并进行预测模型的构建,其R2值可达86.32%,平均绝对误差(MAE)为2.11±1.81 d。结论 本实验基于埋葬大鼠模型所建立的骨骼微生物的死亡时间推断模型具有较好的准确性,可为埋葬案件的死亡时间推断提供参考。

Objective To study the community succession pattern of skeletal microorganisms in buried corpses and to infer postmortem interval(PMI)based on rat models,providing reference for inferring the time of death in burial cases.Methods Twenty rats were buried in soil with a depth of 20 cm,and the upper limb humerus and lower limb femur were collected at five time points(7 d,14 d,30 d,45 d,60 d)respectively.Then the 16S rRNA gene of bacteria was sequenced and the data was analyzed.We constructed a random forest model for predicting the time of death using machine learning algorithm.Results During the 60 days decomposition process of buried rats,there was no significant difference between the microbial communities in the upper limb humerus and those in the lower limb femur,but there were significant differences in microbial communities among different time points.In the process of putrefaction,the relative abundance of Firmicutes and Bacteroidetes in the skeleton showed an overall upward trend,while the relative abundance of Proteobacteria and Actinobacteria showed a downward trend.The random forest algorithm was used to screen 89 OTUs related to PMI as markers and to build a prediction model.The R^(2)value was 86.32%,and the average absolute error(MAE)was 2.11±1.81 days.Conclusion The death time inference model based on the skeletal microorganisms of rat models in this experiment has a good accuracy and can provide reference for burial cases.