Microbiome association analyses are complicated by the fact that microbiomes are very specific to the individual and that many microbes are functionally redundant. To tackle these challenges, we have developed the functional species group (FSG) association analysis, where we associate groups of species with overlapping functional, metabolic or ecological potentials, rather than individual species, to e.g. a clinical measure. Our MGS’ have been annotated with functions and functional modules to facilitate functional species group (FSG) analysis.
This approach has in several cases proven very powerful in deriving meaningful associations Pedersen et al. (Nature Protocols, 2018), Hansen et al. (Nature com., 2018), Roager et al., (Gut, 2017), and perhaps most notable in Pedersen et al. (Nature, 2016), where it was used to link gut FSG to serum metabolite levels. Importantly, the FSG associations often comes with a functional description, namely of the functions that define the FSG. As an example, the serum metabolites in Pedersen et al. were linked to metabolic pathways producing the metabolites.