Early postmortem interval (EPMI) estimation using differentially expressed gene transcripts

•Changes in gene transcript levels after death were detected using microarray.•The early PMI was estimated by gene transcript levels using qPCR.•Errors of PMI estimation by RNA random degradation were reduced.•Ternary equations of temperature, PMI and ΔCq were constructed and visualized.•Four genes with higher R2 and smaller errors were found. Genes differentially expressed after death were selected to construct a mathematical model for early postmortem interval estimation. Sprague Dawley rats were sacrificed and placed at temperatures of 4 °C, 15 °C, 25 °C, and 35 °C. Brain tissues were collected at 0, 6, 12, 18, and 24 h after death and total RNA was extracted. Changes in gene transcript levels after death were detected using microarray expression profiling and differentially expressed genes was screened. Expanded experiments were performed to validate gene transcript levels at different temperatures using the reverse transcription real-time quantitative polymerase chain reaction. Six genes with high coefficients of determination were chosen for construction of mathematical models. Optimal ternary cubic equations were built using R software with temperature, postmortem interval and ΔCq defined as the independent variable x, y and z, respectively. Equations were converted into a three-dimensional visual statistical model using MATLAB. Animal samples were used to validate the mathematical models. Results showed that the 5srRNA showed best stability at four temperatures. The genes Ninj2, Grifin, Arpp19, and Hopx showed high coefficients of determination (>80%) and low error (