Currently, the co-occurrence network were built base on correlation, regression, or other statistical methods. However, these results are ambiguous for understand the interactions. Is there some other method to identify the interactions in microbial network?
You pose a good question - because co-occurrence networks are just that - and the assumption that co-occurrence translates to a a real ecological relationship is not necessarily strong. Clues (but not proof) can come from the phylogenetic assignments (e.g., a positive co-occurrence with a dominant phylotype with a known physiology such as N-fixation could be inferred as a positive trophic interaction), but real proof would only come with experimental studies (good, old-fashioned co-culturing experiments, for example).
- Badges
- Science topic