Collaborative Research: MIM: Using multilayer interaction networks to predict microbiome assembly and function

Microbes play important roles in most ecosystems on Earth and contribute to plant, animal, and human well-being. Understanding microbial community structure and function is necessary for our ability to (1) manage human and animal health and plant productivity and (2) predict ecosystem responses to environmental changes. Consequently, there is an urgent need to understand how microbes interact with each other and with the environment (including their plant, animal, or human hosts).

In this project, the researchers will combine mathematical models and data analyses to uncover the mechanisms that govern the structure and dynamics of microbial communities and species interactions. The novelty of their approach is that it integrates models of different complexity with data on microbial dynamics in different host species to guide the construction, exploration, and testing of those models.

The researchers will develop simple models to explicitly describe different positive and negative interactions among small groups of microbes, such as competition for resources, mutually beneficial exchange of chemicals, or inhibition by toxins. They will also develop multispecies, multilayer network models that reflect the complexity of microbial communities and explore their responses to different disturbances (specifically, invasions by pathogens, antibiotic treatments, or resource-level alterations).

The results of this project will improve understanding of how microbial communities function and how they can be modified to improve animal, plant, and human health. Postdoctoral researchers and students will receive interdisciplinary training in microbial ecology, mathematical modeling, and network science. The researchers will make the approaches and the results of the project available through publications, teaching, and presentations to other scientists and the public.

The project combines theoretical frameworks from ecology, mathematics, and physics — including resource-based ecology, multilayer network analysis, and metacommunity theory — to gain a mechanistic understanding of how microbial interactions determine community dynamics and function. The researchers will develop resource-based models of microbial interactions, and use multilayer network theory to integrate the interaction modules in heterogeneous network structures that include multiple relations, multiple subsystems, multiple scales, and time-dependence.

They will use a tensorial formalism to keep track of both intralayer and interlayer interactions. Their models will explore how different microbial interactions and environmental conditions influence community structure, stability, resilience, and function, and test their models with existing data sets on host-associated microbiomes. The project will advance our understanding of microbial community ecology and help address the problems of managing diverse microbiomes for host and/or ecosystem benefits.

Analyses of dynamic multispecies models and multilayer networks will be applicable to other natural and social systems. The researchers will contribute to open-science efforts through depositing the models and the code in public repositories. They will incorporate the project results into both undergraduate and graduate courses, including the MSU–UCLA distributed seminar.

They will also develop inquiry-based educational modules on microbial communities to increase quantitative reasoning skills and microbial-ecology knowledge in K–12 students. Partial funding was provided by the Math Biology program in the MPS Directorate.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.