Simulation of bacteria interaction networks: from topology to species abundances

Background: Microorganisms frequently coexist and establish complex relationships within their environments. Recent advancements in high-throughput 16S rDNA sequencing techniques have significantly improved our capacity to explore the factors that influence bacterial community organization. However, despite the development of numerous inference methods, the lack of a well-established biological truth presents challenges in validating the obtained results. Therefore, in-silico solutions are critical for simulating realistic gold standards. Methods: We introduce N2SIMBA, a modular algorithmic approach that begins with a known weighted and directed network topology, which is assumed to represent the true nature of microbial interactions. This network serves as a foundation to generate the rules of a Microbial Consumer Resource Model (MiCRM) based on the framework established by Goldford et al. (2018). This model describes the bacterial community characterized by these interactions through a set of Ordinary Differential Equations (ODEs). The subsequent sequencing process is simulated using metaSPARSim as described by Patuzzi et al. (2019). Results: N2SIMBA forms the core component of a systematic framework that allows for the assumption of a gold standard bacterial interaction network and enables the simulation of bacterial community evolution driven by these interactions under varying environmental conditions. The results indicate that, in a resource-rich environment, the bacterial community stabilizes to a relative count distribution similar to the topology of the assumed bacteria interaction network, thereby promoting community resilience and growth. Conversely, a resource-poor environment limits the survival of species the survival of species to those that are highly interconnected within the assumed interaction network. Conclusions: N2SIMBA significantly enhances our understanding of bacterial community organization, providing valuable tools for investigating hypotheses in-silico and evaluating methodologies for inferring bacterial interactions.