The gut microbiome plays a pivotal role in the human body, and perturbations in its composition have been linked to various disorders. Stability is an essential property of a healthy human gut microbiome, which allows it to maintain its functional richness under the external influences. This property has been explored through two distinct methodologies - mathematical modelling based on ecological principles and statistical analysis drawn from observations in interventional studies. Here we conducted a meta-analysis aimed to compare the two approaches utilising the data from 9 interventional and time series studies encompassing 3512 gut microbiome profiles obtained via 16S rRNA gene sequencing. By employing the previously published compositional Lotka-Volterra method, we modelled the dynamics of the microbial community and evaluated ecological stability measures. These measures were compared to those based on observed microbiome changes. There was a substantial correlation between the outcomes of the two approaches. Particularly, local stability assessed within the ecological paradigm was positively correlated with observational stability measures accounting for the compositional nature of microbiome data. Additionally, we were able to reproduce the previously reported inverse relationship between the community's robustness to microorganism loss and local stability, attributed to the distinct impacts of coefficient characterising the network decomposition on these two stability assessments. Our findings demonstrate harmonisation between the ecological and observational approaches to microbiome analysis, advancing the understanding of healthy gut microbiome concept. This paves the way to develop efficient microbiome-targeting interventions for disease prevention and treatment.
Keywords: Dietary interventions; Human gut microbiome; Meta-analysis; Microbiome dynamics; Responders; Stability; Time series.
© 2023 The Authors.