Using causality and correlation analysis to decipher microbial interactions in activated sludge

Abstract

Network theory is widely used to understand microbial interactions in activated sludge and numerous other artificial and natural environments. However, when using correlation-based methods, it is not possible to identify the directionality of interactions within microbiota. Based on the classic Granger test of sequencing-based time-series data, a new Microbial Causal Correlation Network (MCCN) was constructed with distributed ecological interaction on the directed, associated links. As a result of applying MCCN to a time series of activated sludge data, we found that the hub species OTU56, classified as belonging the genus Nitrospira, was responsible for completing nitrification in activated sludge, and mainly interacted with Proteobacteria and Bacteroidetes in the form of amensal and commensal relationships, respectively. Phylogenetic tree suggested a mutualistic relationship between Nitrospira and denitrifiers. Zoogloea displayed the highest ncf value within the classified OTUs of the MCCN, indicating that it could be a foundation for activated sludge through forming the characteristic cell aggregate matrices into which other organisms embed during floc formation. Overall, the introduction of causality analysis greatly expands the ability of a network to shed a light on understanding the interactions between members of a microbial community.