With the advance of Next Generation Sequencing, more and more detailed studies of microbial communities are being available. The interest in the role of microbiome in various aspects of human life is also on the rise. As such, understanding the microbial community dynamics has never been more important.
As a way of understanding microbial community dynamics, Microbial Interaction Networks (MIN) have been inferred with various statistical and model based methods through the analysis of microbial abundance profiles. MINs have often been modelled and interpreted with various flavours of Lotka-Volterra equations.
Microbial interactions have been identified to be varying in nature. Mutualism, competition, parasitism etc. are some ways of describing them. However all these interactions, and the microbial interactions networks are assumed to be static in existing work. A potential reason is that, the dynamics of microbial interactions are not verifiable through in-vitro means. Our attempt is in exploring these interactions through the lens of time-series data analysis.
We explore the world of microbial interactions where the individual interactions are not studied under the assumptions of statis. In our exploration we provide evidence for the dynamic nature of these interactions, compare and contrast how various interactions vary with time. We attempt to model the interaction dynamics over short and long time periods.
We present our findings applied to simulated data sets, as well as to human microbiome samples.
In conclusion we highlight the importance of considering microbial interactions as dynamic systems, and the need for models which account for the dynamic nature of these interactions.