Testing the stress gradient hypothesis in soil bacterial communities associated with vegetation belts in the Andean Atacama Desert

BACKGROUND

Soil microorganisms are in constant interaction with plants, and these interactions shape the composition of soil bacterial communities by modifying their environment. However, little is known about the relationship between microorganisms and native plants present in extreme environments that are not affected by human intervention. Using high-throughput sequencing in combination with random forest and co-occurrence network analyses, we compared soil bacterial communities inhabiting the rhizosphere surrounding soil (RSS) and the corresponding bulk soil (BS) of 21 native plant species organized into three vegetation belts along the altitudinal gradient (2400-4500 m a.s.l.) of the Talabre-Lejía transect (TLT) in the slopes of the Andes in the Atacama Desert. We assessed how each plant community influenced the taxa, potential functions, and ecological interactions of the soil bacterial communities in this extreme natural ecosystem. We tested the ability of the stress gradient hypothesis, which predicts that positive species interactions become increasingly important as stressful conditions increase, to explain the interactions among members of TLT soil microbial communities.

RESULTS

Our comparison of RSS and BS compartments along the TLT provided evidence of plant-specific microbial community composition in the RSS and showed that bacterial communities modify their ecological interactions, in particular, their positive:negative connection ratios in the presence of plant roots at each vegetation belt. We also identified the taxa driving the transition of the BS to the RSS, which appear to be indicators of key host-microbial relationships in the rhizosphere of plants in response to different abiotic conditions. Finally, the potential functions of the bacterial communities also diverge between the BS and the RSS compartments, particularly in the extreme and harshest belts of the TLT.

CONCLUSIONS

In this study, we identified taxa of bacterial communities that establish species-specific relationships with native plants and showed that over a gradient of changing abiotic conditions, these relationships may also be plant community specific. These findings also reveal that the interactions among members of the soil microbial communities do not support the stress gradient hypothesis. However, through the RSS compartment, each plant community appears to moderate the abiotic stress gradient and increase the efficiency of the soil microbial community, suggesting that positive interactions may be context dependent.

Enhancing Microbiome Research through Genome-Scale Metabolic Modeling

Construction and analysis of genome-scale metabolic models (GEMs) is a well-established systems biology approach that can be used to predict metabolic and growth phenotypes. The ability of GEMs to produce mechanistic insight into microbial ecological processes makes them appealing tools that can open a range of exciting opportunities in microbiome research. Here, we briefly outline these opportunities, present current rate-limiting challenges for the trustworthy application of GEMs to microbiome research, and suggest approaches for moving the field forward.

Predicting microbiomes through a deep latent space

Motivation

Microbial communities influence their environment by modifying the availability of compounds, such as nutrients or chemical elicitors. Knowing the microbial composition of a site is therefore relevant to improve productivity or health. However, sequencing facilities are not always available, or may be prohibitively expensive in some cases. Thus, it would be desirable to computationally predict the microbial composition from more accessible, easily-measured features.

Results

Integrating deep learning techniques with microbiome data, we propose an artificial neural network architecture based on heterogeneous autoencoders to condense the long vector of microbial abundance values into a deep latent space representation. Then, we design a model to predict the deep latent space and, consequently, to predict the complete microbial composition using environmental features as input. The performance of our system is examined using the rhizosphere microbiome of Maize. We reconstruct the microbial composition (717 taxa) from the deep latent space (10 values) with high fidelity (>0.9 Pearson correlation). We then successfully predict microbial composition from environmental variables, such as plant age, temperature or precipitation (0.73 Pearson correlation, 0.42 Bray–Curtis). We extend this to predict microbiome composition under hypothetical scenarios, such as future climate change conditions. Finally, via transfer learning, we predict microbial composition in a distinct scenario with only 100 sequences, and distinct environmental features. We propose that our deep latent space may assist microbiome-engineering strategies when technical or financial resources are limited, through predicting current or future microbiome compositions.

Availability and implementation

Software, results and data are available at https://github.com/jorgemf/DeepLatentMicrobiome

Supplementary information

Supplementary data are available at Bioinformatics online.

Metabolic modelling approaches for describing and engineering microbial communities

Microbes do not live in isolation but in microbial communities. The relevance of microbial communities is increasing due to growing awareness of their influence on a huge number of environmental, health and industrial processes. Hence, being able to control and engineer the output of both natural and synthetic communities would be of great interest. However, most of the available methods and biotechnological applications involving microorganisms, both in vivo and in silico, have been developed in the context of isolated microbes. In vivo microbial consortia development is extremely difficult and costly because it implies replicating suitable environments in the wet-lab. Computational approaches are thus a good, cost-effective alternative to study microbial communities, mainly via descriptive modelling, but also via engineering modelling. In this review we provide a detailed compilation of examples of engineered microbial communities and a comprehensive, historical revision of available computational metabolic modelling methods to better understand, and rationally engineer wild and synthetic microbial communities.

Data-driven classification of the certainty of scholarly assertions

The grammatical structures scholars use to express their assertions are intended to convey various degrees of certainty or speculation. Prior studies have suggested a variety of categorization systems for scholarly certainty; however, these have not been objectively tested for their validity, particularly with respect to representing the interpretation by the reader, rather than the intention of the author. In this study, we use a series of questionnaires to determine how researchers classify various scholarly assertions, using three distinct certainty classification systems. We find that there are three distinct categories of certainty along a spectrum from high to low. We show that these categories can be detected in an automated manner, using a machine learning model, with a cross-validation accuracy of 89.2% relative to an author-annotated corpus, and 82.2% accuracy against a publicly-annotated corpus. This finding provides an opportunity for contextual metadata related to certainty to be captured as a part of text-mining pipelines, which currently miss these subtle linguistic cues. We provide an exemplar machine-accessible representation-a Nanopublication-where certainty category is embedded as metadata in a formal, ontology-based manner within text-mined scholarly assertions.

FLYCOP: metabolic modeling-based analysis and engineering microbial communities

Motivation

Synthetic microbial communities begin to be considered as promising multicellular biocatalysts having a large potential to replace engineered single strains in biotechnology applications, in pharmaceutical, chemical and living architecture sectors. In contrast to single strain engineering, the effective and high-throughput analysis and engineering of microbial consortia face the lack of knowledge, tools and well-defined workflows. This manuscript contributes to fill this important gap with a framework, called FLYCOP (FLexible sYnthetic Consortium OPtimization), which contributes to microbial consortia modeling and engineering, while improving the knowledge about how these communities work. FLYCOP selects the best consortium configuration to optimize a given goal, among multiple and diverse configurations, in a flexible way, taking temporal changes in metabolite concentrations into account.

Results

In contrast to previous systems optimizing microbial consortia, FLYCOP has novel characteristics to face up to new problems, to represent additional features and to analyze events influencing the consortia behavior. In this manuscript, FLYCOP optimizes a Synechococcus elongatus-Pseudomonas putida consortium to produce the maximum amount of bio-plastic (PHA, polyhydroxyalkanoate), and highlights the influence of metabolites exchange dynamics in a four auxotrophic Escherichia coli consortium with parallel growth. FLYCOP can also provide an explanation about biological evolution driving evolutionary engineering endeavors by describing why and how heterogeneous populations emerge from monoclonal ones.

Availability and implementation

Code reproducing the study cases described in this manuscript are available on-line: https://github.com/beatrizgj/FLYCOP.

Supplementary information

Supplementary data are available at Bioinformatics online.

Robust and automatic definition of microbiome states

Analysis of microbiome dynamics would allow elucidation of patterns within microbial community evolution under a variety of biologically or economically important circumstances; however, this is currently hampered in part by the lack of rigorous, formal, yet generally-applicable approaches to discerning distinct configurations of complex microbial populations. Clustering approaches to define microbiome “community state-types” at a population-scale are widely used, though not yet standardized. Similarly, distinct variations within a state-type are well documented, but there is no rigorous approach to discriminating these more subtle variations in community structure. Finally, intra-individual variations with even fewer differences will likely be found in, for example, longitudinal data, and will correlate with important features such as sickness versus health. We propose an automated, generic, objective, domain-independent, and internally-validating procedure to define statistically distinct microbiome states within datasets containing any degree of phylotypic diversity. Robustness of state identification is objectively established by a combination of diverse techniques for stable cluster verification. To demonstrate the efficacy of our approach in detecting discreet states even in datasets containing highly similar bacterial communities, and to demonstrate the broad applicability of our method, we reuse eight distinct longitudinal microbiome datasets from a variety of ecological niches and species. We also demonstrate our algorithm’s flexibility by providing it distinct taxa subsets as clustering input, demonstrating that it operates on filtered or unfiltered data, and at a range of different taxonomic levels. The final output is a set of robustly defined states which can then be used as general biomarkers for a wide variety of downstream purposes such as association with disease, monitoring response to intervention, or identifying optimally performant populations.

MDPbiome: microbiome engineering through prescriptive perturbations

Motivation

Recent microbiome dynamics studies highlight the current inability to predict the effects of external perturbations on complex microbial populations. To do so would be particularly advantageous in fields such as medicine, bioremediation or industrial scenarios.

Results

MDPbiome statistically models longitudinal metagenomics samples undergoing perturbations as a Markov Decision Process (MDP). Given a starting microbial composition, our MDPbiome system suggests the sequence of external perturbation(s) that will engineer that microbiome to a goal state, for example, a healthier or more performant composition. It also estimates intermediate microbiome states along the path, thus making it possible to avoid particularly undesirable/unhealthy states. We demonstrate MDPbiome performance over three real and distinct datasets, proving its flexibility, and the reliability and universality of its output 'optimal perturbation policy'. For example, an MDP created using a vaginal microbiome time series, with a goal of recovering from bacterial vaginosis, suggested avoidance of perturbations such as lubricants or sex toys; while another MDP provided a quantitative explanation for why salmonella vaccine accelerates gut microbiome maturation in chicks. This novel analytical approach has clear applications in medicine, where it could suggest low-impact clinical interventions that will lead to achievement or maintenance of a healthy microbial population, or alternately, the sequence of interventions necessary to avoid strongly negative microbiome states.

Availability and implementation

Code (https://github.com/beatrizgj/MDPbiome) and result files (https://tomdelarosa.shinyapps.io/MDPbiome/) are available online.

Supplementary information

Supplementary data are available at Bioinformatics online.

Predicting Protein Relationships to Human Pathways through a Relational Learning Approach Based on Simple Sequence Features

Biological pathways are important elements of systems biology and in the past decade, an increasing number of pathway databases have been set up to document the growing understanding of complex cellular processes. Although more genome-sequence data are becoming available, a large fraction of it remains functionally uncharacterized. Thus, it is important to be able to predict the mapping of poorly annotated proteins to original pathway models.

RESULTS

We have developed a Relational Learning-based Extension (RLE) system to investigate pathway membership through a function prediction approach that mainly relies on combinations of simple properties attributed to each protein. RLE searches for proteins with molecular similarities to specific pathway components. Using RLE, we associated 383 uncharacterized proteins to 28 pre-defined human Reactome pathways, demonstrating relative confidence after proper evaluation. Indeed, in specific cases manual inspection of the database annotations and the related literature supported the proposed classifications. Examples of possible additional components of the Electron transport system, Telomere maintenance and Integrin cell surface interactions pathways are discussed in detail.

AVAILABILITY

All the human predicted proteins in the 2009 and 2012 releases 30 and 40 of Reactome are available at http://rle.bioinfo.cnio.es.