Abstract
BACKGROUND: Timely and accurate pregnancy detection in the dairy animals is essential for minimizing calving intervals to accelerate reproductive turnover and increase milk production from livestock farming. Urine, a non-invasive sample, is being investigated in many studies as it serves as a pool depicting the varied array of metabolites associated with intricate physiological changes. The study thus aimed to analyze GC-MS based urinary metabolites of pregnant (n = 13) and non-pregnant (n = 13) buffaloes using Bovine Metabolome Database. RESULTS: A total of 102 unique metabolites were annotated using acetonitrile, dichloromethane, hexane and methanol solvents for metabolite extraction. Forty metabolites including Ethanolamine, aminomalonic acid, myoinositol, cellobiose, phenylacetic acid and phenol showed a significant fold change in pregnant animals, samples. Organonitrogen compounds and organic carbonic acids and derivatives chemical classes were in higher abundance in pregnant buffaloes. Three metabolites viz. 3-hydroxybutyric acid, phenol and ethanolamine were identified as potential biomarkers with AUC values of 1. Supervised multivariate analysis by oPLSDA showed good separation of groups with R2Y and Q2 values of 0.958 and 0.861 respectively. Glycerophospholipid, amino acid and nucleotide metabolism associated metabolites were observed in enrichment analysis. Network analysis of significant metabolites further showed a dense network of pregnancy associated metabolic reprogramming of carbohydrate and energy releasing pathways, amino acid and nitrogen metabolism pathways and lipid metabolism and signaling pathways. CONCLUSION: The study provides insights into allied metabolic reshuffling as observed by non-invasive urinary metabolic profile of buffaloes during pregnancy and identified potential urine based biomarkers associated with pregnancy. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12917-026-05427-6.