Ecological network reconstruction

In studying Ecological network or Microbiome, we often don’t know the interaction, but only have the abundance profile across species and samples (or time). Can we infer the interaction network from it? This is a Network reconstruction problem (or more specifically Network reconstruction from dynamics).

Xiao2017mapping provides a method that doesn’t rely on particular population dynamics. Young2019reconstruction, Peixoto2019network, Hoffmann2020community, and other Network science + Bayesian inference approaches may be useful.

Lo2017inferring proposes a method to leverage existing biological knowledge.

Fricke2022collapse uses Deep learning to reconstruct historical ecological networks.