Uncovering Interactions between Plant Metabolism and Plant-Associated Bacteria in Huanglongbing-Affected Citrus Cultivars Using Multiomics Analysis and Machine Learning

Huanglongbing (HLB) is a highly destructive disease that inflicts significant economic losses on the citrus industry worldwide but with no cure available. However, microbiomes formulated by citrus plants may serve as disease antagonists, increasing the level of HLB tolerance. This study established an integrated analysis of untargeted metabolomics and microbiomics data for different citrus cultivars, providing critical insights into the interactions between plant metabolism and plant-associated bacteria in the development of HLB. Machine learning models were applied to screen important metabolites and bacteria in multiple citrus materials, and the selected metabolites were then analyzed to identify essential pathways enriched in the plant and to correlate with the selected bacteria. Results demonstrated that the regulation of plant pathways, especially ABC transporters and ubiquinone and other terpene-ubiquinone biosynthesis pathways, could affect the microbial community structure, indicating potential solutions for controlling HLB by modulating bacteria in citrus plants or breeding tolerant citrus cultivars.

Automated summary

This study utilizes an integrated analysis of untargeted metabolomics and microbiomics data to explore the interactions between plant metabolism and plant-associated bacteria in the development of HLB. Machine learning models are applied to identify important metabolites and bacteria and provide potential solutions for controlling HLB.