Master Thesis: Assessing Evidence for Higher-Order Interactions in Microbial Time-Series Data
Reconstructing ecological interaction networks from empirical data remains a major challenge. Here, we incorporate higher-order interactions (HOIs) into networks inferred from experimental ecological time series. We first fit a pairwise generalized Lotka–Volterra (gLV) model and then extend it to include interactions among three species, yielding a triadic gLV model. The inferred networks are compared with two null models: one based on link redistribution and another based on temporal randomization before inference.
Pairwise networks are less dense than expected under the null models and contain a higher proportion of negative interactions. By contrast, networks including HOIs are denser than expected and contain many triadic interactions, although pairwise interactions remain predominant. We also find evidence of simplicial closure: triadic interactions supported by only two underlying pairwise links are underrepresented, whereas those supported by all three are overrepresented. Degree and hyperdegree are correlated, a pattern partly reproduced by the temporal-randomization null model. These results provide insight into the higher-order organization of ecological networks and motivate further experimental investigation of the signs and mechanisms of such interactions.
Supervisor: Juan Fernández Gracia
Committee: Tomás Sintes, Víctor M. Eguíluz, Juan Fernández Gracia
This Master Thesis will be broadcasted in the following zoom link: https://us06web.zoom.us/j/89466064429?pwd=po9p99eAEYVPaNI8xIIGoOIz0hOqaF.1
Detalls de contacte:
Juan Fernández Gracia Contact form