Boolean analysis reveals systematic interactions among low-abundance species in the human gut microbiome

Abstract

The analysis of microbiome compositions in the human gut has gained increasing interest due to the broader availability of data and functional databases and substantial progress in data analysis methods, but also due to the high relevance of the microbiome in human health and disease. While most analyses infer interactions among highly abundant species, the large number of low-abundance species has received less attention. Here we present a novel analysis method based on Boolean operations applied to microbial co-occurrence patterns. We calibrate our approach with simulated data based on a dynamical Boolean network model from which we interpret the statistics of attractor states as a theoretical proxy for microbiome composition. We show that for given fractions of synergistic and competitive interactions in the model our Boolean abundance analysis can reliably detect these interactions. Analyzing a novel data set of 822 microbiome compositions of the human gut, we find a large number of highly significant synergistic interactions among these low-abundance species, forming a connected network, and a few isolated competitive interactions.

The microbiome and cancer

“…The individual human microbiome appears to be relatively persistent, although it is occasionally dramatically perturbed by antibiotic treatment [8]; however, it should be noted that very little study has been made of the effects of deliberate changes in diet. Certainly, much of the microbiome can be repopulated by organisms protected in crypts within the epithelial mucous layer [16], and interactions between less abundant species may be important for maintenance of the overall structure of the individual microbiome [19]. Studies examining the general association of colon cancer with the microbiome (see below) have yielded varying results, depending on the population examined, but an underpinning theme is that microbiota associated with a plant diet tend to be the ones associated with a lower incidence of colon cancer.…”

Shrinkage improves estimation of microbial associations under different normalization methods

“…However, transformations such as asinh and clr may be preferred since they are faster to compute than vst, while providing similar statistical properties. The resulting shrinkage correlation estimates can then also serve as input for more involved direct microbial network inference workflows b a that account for transitive correlations, adjust for additional covariates, or model latent effects (12,37,47,48).…”

“…Multiple studies have found a higher prevalence of positive associations between taxonomically related taxa in human gut datasets (12,(37)(38)(39)(40). We thus use taxonomic coherence, measured by assortativity (41), as independent summary statistic for relevance networks.…”

Akkermansia and Microbial Degradation of Mucus in Cats and Dogs: Implications to the Growing Worldwide Epidemic of Pet Obesity

“…On the other extreme, we have the rare and less prevalent microbes, which have been detected only in a few individuals and in very low abundance or not at all. Unfortunately, we know little about the contribution of these low abundant groups to overall gut homeostasis [179]; in fact, some researchers suggest and promote the removal of low abundant microbes from analysis in order to detect "true" phylotypes [180]. This is not a trivial topic, at least in other environments such as soil it has been shown that the low-abundance bacteria play a fundamental biological role [181] and may, in fact, be keystone species regulating the function of different microbial environments, including host-associated microbiomes [182].…”