Table 3.

Mathematical models used for microbiome engineering.

Approach	Model	Main model input	Feature	References
Bottom‐up engineering	Lotka–Volterra pairwise model (LVPM)	Pairwise interactions	Based on pairwise interactions	[32, 47]
	Lotka‐Volterra mechanistic model (LVMM)	Interaction mechanisms	The model structure may be more complex than LVPM but can be simplified or solved by simulations	[33, 54, 89]
	Genome‐scale metabolic model (GSMM)	Genomic data	Link the metabolic traits of a single member with the properties of the community	[31, 61, 125, 126]
	Individual‐based model (IBM)	Interaction mechanisms; spatial position	Simulate the dynamics of a community in spatially structured environments	[76, 127, 128, 129]
	Computaion of microbial ecosystems (COMETS)	Genomic data; spatial position	Combine features of IBM and GSMM	[130, 131]
Top‐down engineering	Consumer‐resource model (CRM) integrated with ecological interactions	Parameters regarding nutrient availability, interactions, etc.	Predict the dynamics of the enriched multispecies community	[98, 107]
	IBM	Parameters regarding nutrient availability, interactions, etc.	Capture the dynamics of the multispecies community with artificial selection	[105]