Inferring microbial interaction network from microbiome data using RMN algorithm

Abstract

BackgroundMicrobial interactions are ubiquitous in nature. Recently, many similarity-based approaches have been developed to study the interaction in microbial ecosystems. These approaches can only explain the non-directional interactions yet a more complete view on how microbes regulate each other remains elusive. In addition, the strength of microbial interactions is difficult to be quantified by only using correlation analysis.ResultsIn this study, a rule-based microbial network (RMN) algorithm, which integrates regulatory OTU-triplet model with parametric weighting function, is being developed to construct microbial regulatory networks. The RMN algorithm not only can extrapolate the cooperative and competitive relationships between microbes, but also can infer the direction of such interactions. In addition, RMN algorithm can theoretically characterize the regulatory relationship composed of microbial pairs with low correlation coefficient in microbial networks. Our results suggested that Bifidobacterium, Streptococcus, Clostridium XI, and Bacteroides are essential for causing abundance changes of Veillonella in gut microbiome. Furthermore, we inferred some possible microbial interactions, including the competitive relationship between Veillonella and Bacteroides, and the cooperative relationship between Veillonella and Clostridium XI.ConclusionsThe RMN algorithm provides the reconstruction of gut microbe networks, and can shed light on the dynamical interactions of microbes in the infant intestinal tract.Electronic supplementary materialThe online version of this article (doi:10.1186/s12918-015-0199-2) contains supplementary material, which is available to authorized users.

Competition and niche separation of pelagic bacteria in freshwater habitats

“…However, while it is intuitively attractive to synoptically assess the relationships between all members of lacustrine microbial assemblages at once (Tsai et al, 2015), the so revealed statistical correlations are causally often rather ambiguous. A conspicuous temporal cooccurrence of genotypes may indicate both, fierce competition or the lack thereof (total niche separation), but might also be due to dispersal patterns within larger metacommunity systems (Crump et al, 2007;Lindstr€ om and Langenheder, 2012), for example, influx from the catchment (Ruiz-Gonz alez et al, 2015;Niño-Garcia et al, 2016a).…”

MetaMIS: a metagenomic microbial interaction simulator based on microbial community profiles

“…The conventional approach is to observe the growth behavior in mixed cultures of only a very few microorganisms [8]. Recently, high-throughput interaction inference approaches, such as Sparse Correlations for Compositional data (SparCC) [9], the Learning Interactions from MIcrobial Time Series (LIMITS) algorithm [10], co-occurrence networks [11], the SParse InversE Covariance estimation for Ecological ASsociation Inference (SPIEC-EASI) [12], and the Rule-based Microbial Network (RMN) algorithm [13], have been proposed for modeling microscale dynamics using 16S rRNA marker gene sequences. These approaches may be roughly divided into two categories.…”

“…Although correlation is straightforward and easy to conduct, it nevertheless does not seem to be a proper measure of species interactions, and is limited to inferring non-directional interactions [11, 12]. Modeling-centered approaches, on the other hand, including the LIMITS [10], SPIEC-EASI [12], and RMN [13] algorithms, rest on special biological assumptions and statistical techniques, and usually employ a combined strategy in order to infer microbial interactions. LIMITS, for instance, combines a spare linear regression with a bootstrapping strategy in order to incorporate interactive relations iteratively into an interaction network [10].…”

“…SPIEC-EASI assumes the underlying ecological association network to be sparse and accordingly relies on sparse inverse covariance selection and a neighborhood selection strategy to reconstruct a non-directional interaction network [12]. The RMN algorithm bypasses the NP-hard problem of finding a network with the optimum number of interactions and proceeds directly to the construction of a triplet subnetwork in which the triplet has a convergent recipient that is repressed by one interaction and simultaneously activated by another [13]. …”

Massively parallel screening of synthetic microbial communities

“…The complexity of microbial interactions and environmental dependencies (13)(14)(15)(16) can lead to unpredictable behaviors even in apparently simple communities, posing a challenge to design of consortia. Addressing this challenge likely requires the integration of multiple approaches, including the reverse-engineering of natural communities (e.g., inference-based co-occurrence analysis) (17) and further development of forward-engineering strategies (e.g., metabolic flux-balance analysis) (18,19). Complimentarily, screening experimentally constructed synthetic combinations of strains can identify consortia with desired properties or validate rational designs (20)(21)(22).…”