Exploring Computational Inference of Microbial Interactions and their Dynamics

Abstract

Next Generation Sequencing has increased detailed studies of microbial communities. Understanding microbial interactions is crucial for understanding these communities. Microbial Interaction Networks (MINs) have been inferred using microbial abundance profiles, using multiple methods and frameworks. In this paper we explore current approaches in inferring microbial interactions and demonstrate that they are dynamic in nature, and propose a visualisation approach for observing dynamic nature of MINs. Our work highlights that even very rare microbiota have bursts of high abundance at some time points. Further, we suggest possible improvements in the field of computational inference of MINs.