Community detection in a plant-based fermentation knowledge graph

Overview & Synopsis

Microbial data have rapidly expanded across diverse resources, including PubMed, omics datasets, and strain collections, creating a need for integration and standardization. To address this, Omnicrobe was developed, a relational database that regroups heterogeneous microbial information. From this resource, we extracted plant-based fermentationrelated data and represented them as a knowledge graph. This study investigates whether the structure of this graph contains sufficient signals to generate new biological hypotheses about fermentation processes. To evaluate this, we detect communities within the graph and perform a first evaluation of their characteristics.