论文部分内容阅读
A metabonomics method was developed to investigate the physiological effect of electric induced stress in female Sprague-Dawley (SD) rats using UPLC/QTOF MS.Artificial neural networks (ANNs) and neurofuzzy logic software (Intelligensys) were applied for data mining activities to identify potential biomarkers in the database.A database with 144 data records and 4000 independent variables (biochemical substance) was collected from a designed animal study investigating the variation of the composition of chemical substances in biological samples as a result of the application of electrically induced stimulation over a 6-days period.With principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) of urinary metabolites, clear differences between urine samples collected from rats electriclly induced in different days were observed and the trajectory of the metabolic profiles exhibited a tendency recovering to the control rats by treating Baixiangdan capsule.Using the data available, a neurofuzzy logic model was successfully generated which enabled the discovery of 14 potential biomarkers representing major cause-effect relationships between the composition of biochemical substance and the development of disease.And the result showed high overlap with PLS-DA model.The predictability of both ANN models is similar with the validation R2 above 0.97, indicating that the 14 biomarkers explain 97% of variability of the cause-effect relationship between biochemical substance and the development of disease.Thus these are considered to be the likely bio-markers associated with induced stress in this study.Furthermore, a distinct difference in serum metabolic profiling was achieved in control, induced stress and treatment groups.Metabolites which were important for classification were identified based on their exact mass measurement from UPLC/QTOF-MS.Amino acid metabolites such as 2-aminoadipate, 4-hydroxyglutamate, hippuric acid, spermine, phenylalanine and iysine were identified in both urine and serum.Moreover, some significantly changed metabolites in treatment group were found and indicated that Baixiangdan capsule may play its role by suppressing glutamine metabolism, inhibiting the activity of glutamate, up-regulating the level of estrogen and inhibiting the increasing of neurotransmitter in serum.The methodology developed in this study can be improved understanding of the development of electrically induced stress in female Sprague-Dawley (SD) rats and potentially applied to broader applications in the pharmaceutical and medical sciences.