A community metabolic model predicts that species A cross-feeds species B by producing a metabolite that B requires but cannot synthesize. If species A is eliminated (e.g., by an antibiotic), what does the model predict?
ANothing changes, because microbial communities are always stable
BSpecies B declines or disappears unless another community member can produce the same metabolite, and species that depended on B may also be affected — creating a cascade through the community's metabolic interaction network
CSpecies B immediately evolves the ability to produce the metabolite
DSpecies B switches to a different carbon source with no fitness cost
Cross-feeding dependencies create ecological fragility — removing a key metabolite producer can trigger cascading effects through the community. Community metabolic models (like those built with MICOM or SteadyCom) explicitly map these metabolite exchange networks and can predict which species are vulnerable to the loss of key producers. The cascade effect explains why narrow-spectrum antibiotics can have community-wide consequences and why microbiome disturbances can be difficult to reverse. Real communities have some redundancy (multiple species producing the same metabolite), but the degree of redundancy varies and is itself a prediction of the model.
Question 2 True / False
Generalized Lotka-Volterra models of microbial communities assume that all interactions between species are mediated by direct cell-cell contact.
TTrue
FFalse
Answer: False
Generalized Lotka-Volterra (gLV) models represent species interactions as pairwise interaction coefficients in a system of ODEs — they capture the net effect of all interaction mechanisms (competition for shared resources, metabolic cross-feeding, toxin production, pH modification) in a single coefficient per species pair. The models are agnostic about mechanism; they describe phenomenological interactions inferred from co-occurrence patterns or perturbation experiments. This is both a strength (computational tractability) and a limitation (the mechanistic basis of interactions is hidden, and interaction coefficients may not be constant across environments).
Question 3 Short Answer
Why are community-level metabolic models (multi-species FBA) more informative than simply summing the individual metabolic models of each species?
Think about your answer, then reveal below.
Model answer: Individual species models assume each organism grows independently with access to all available nutrients. In reality, community members compete for shared substrates (one species consuming a metabolite makes it unavailable to others) and cross-feed (one species' waste product is another's essential nutrient). Community metabolic models explicitly represent these inter-species metabolic interactions, predicting which metabolite exchanges are thermodynamically and stoichiometrically feasible. These emergent community-level behaviors — the mutualistic loops, competitive exclusion patterns, and metabolic division of labor — cannot be predicted by summing individual models because they arise from the interactions between species.
Tools like MICOM, SteadyCom, and OptCom implement community-level FBA by coupling individual species metabolic models through a shared extracellular metabolite pool, allowing species to trade metabolites subject to mass balance and thermodynamic constraints. The predicted metabolite exchanges often match experimental measurements of cross-feeding relationships.