Classification of soybeans from different habitats based on metabolomic–transcriptomic integration

Abstract

AbstractSoybeans are a significant agricultural product in China, with certain geographical locations often yielding higher quality, and thus more expensive, soybean crops. In this study, metabolomics and transcriptomics analyses were conducted on soybean samples from nine regions in Heilongjiang and Liaoning Provinces using untargeted liquid chromatography–mass spectrometry (LC–MS) and Illumina sequencing technologies. The primary objective was to devise an effective and unbiased method for determining the geographical origin of each soybean variety to mitigate potential fraudulent practices. Through multidimensional and unidimensional analyses, successful identification of differentially expressed metabolites (DEMs) and differentially expressed genes (DEGs) was achieved, yielding statistically significant outcomes. Integration of the metabolomics and transcriptomics datasets facilitated the construction of a correlation network model capable of distinguishing soybeans originating from different geographical locations, leading to the identification of significant biomarkers exemplifying noteworthy distinctions. To validate the feasibility of this method in practical applications, partial least squares discriminant analysis was employed to differentiate soybean samples from the nine regions. The results convincingly showcased the applicability and reliability of this approach in accurately pinpointing the geographical origin of soybeans. Distinguishing itself from prior research in soybean traceability, this study incorporates an integrated analysis of metabolomics and transcriptomics data, thereby unveiling biomarkers that offer a more precise differentiation of soybean traits across distinct regions, thereby bridging a critical research gap within the soybean traceability domain. This innovative dual-data integration analysis methodology is poised to enhance the accuracy of soybean traceability tools and lay a new foundation for future agricultural product identification research.