Method development for cross-study microbiome data mining: Challenges and opportunities

On the limits of 16S rRNA gene-based metagenome prediction and functional profiling

Molecular profiling techniques such as metagenomics, metatranscriptomics or metabolomics offer important insights into the functional diversity of the microbiome. In contrast, 16S rRNA gene sequencing, a widespread and cost-effective technique to measure microbial diversity, only allows for indirect estimation of microbial function. To mitigate this, tools such as PICRUSt2, Tax4Fun2, PanFP and MetGEM infer functional profiles from 16S rRNA gene sequencing data using different algorithms. Prior studies have cast doubts on the quality of these predictions, motivating us to systematically evaluate these tools using matched 16S rRNA gene sequencing, metagenomic datasets, and simulated data. Our contribution is threefold: (i) using simulated data, we investigate if technical biases could explain the discordance between inferred and expected results; (ii) considering human cohorts for type two diabetes, colorectal cancer and obesity, we test if health-related differential abundance measures of functional categories are concordant between 16S rRNA gene-inferred and metagenome-derived profiles and; (iii) since 16S rRNA gene copy number is an important confounder in functional profiles inference, we investigate if a customised copy number normalisation with the rrnDB database could improve the results. Our results show that 16S rRNA gene-based functional inference tools generally do not have the necessary sensitivity to delineate health-related functional changes in the microbiome and should thus be used with care. Furthermore, we outline important differences in the individual tools tested and offer recommendations for tool selection.

Exercise and microbiome: From big data to therapy

Exercise is a vital component in maintaining optimal health and serves as a prospective therapeutic intervention for various diseases. The human microbiome, comprised of trillions of microorganisms, plays a crucial role in overall health. Given the advancements in microbiome research, substantial databases have been created to decipher the functionality and mechanisms of the microbiome in health and disease contexts. This review presents an initial overview of microbiomics development and related databases, followed by an in-depth description of the multi-omics technologies for microbiome. It subsequently synthesizes the research pertaining to exercise-induced modifications of the microbiome and diseases that impact the microbiome. Finally, it highlights the potential therapeutic implications of an exercise-modulated microbiome in intestinal disease, obesity and diabetes, cardiovascular disease, and immune/inflammation-related diseases.

More Than the Sum of Its Parts: Unlocking the Power of Network Structure for Understanding Organization and Function in Microbiomes

Plant and soil microbiomes are integral to the health and productivity of plants and ecosystems, yet researchers struggle to identify microbiome characteristics important for providing beneficial outcomes. Network analysis offers a shift in analytical framework beyond "who is present" to the organization or patterns of coexistence between microbes within the microbiome. Because microbial phenotypes are often significantly impacted by coexisting populations, patterns of coexistence within microbiomes are likely to be especially important in predicting functional outcomes. Here, we provide an overview of the how and why of network analysis in microbiome research, highlighting the ways in which network analyses have provided novel insights into microbiome organization and functional capacities, the diverse network roles of different microbial populations, and the eco-evolutionary dynamics of plant and soil microbiomes.

In situ reductive dehalogenation of groundwater driven by innovative organic carbon source materials: Insights into the organohalide-respiratory electron transport chain

In situ bioremediation using organohalide-respiring bacteria (OHRB) is a prospective method for the removal of persistent halogenated organic pollutants from groundwater, as OHRB can utilize H₂ or organic compounds produced by carbon source materials as electron donors for cell growth through organohalide respiration. However, few previous studies have determined the suitability of different carbon source materials to the metabolic mechanism of reductive dehalogenation from the perspective of electron transfer. The focus of this critical review was to reveal the interactions and relationships between carbon source materials and functional microbes, in terms of the electron transfer mechanism. Furthermore, this review illustrates some innovative strategies that have used the physiological characteristics of OHRB to guide the optimization of carbon source materials, improving the abundance of indigenous dehalogenated bacteria and enhancing electron transfer efficiency. Finally, it is proposed that future research should combine multi-omics analysis with machine learning (ML) to guide the design of effective carbon source materials and optimize current dehalogenation bioremediation strategies to reduce the cost and footprint of practical groundwater bioremediation applications.

Machine Learning Techniques for Antimicrobial Resistance Prediction of Pseudomonas Aeruginosa from Whole Genome Sequence Data

AIM

Due to the growing availability of genomic datasets, machine learning models have shown impressive diagnostic potential in identifying emerging and reemerging pathogens. This study aims to use machine learning techniques to develop and compare a model for predicting bacterial resistance to a panel of 12 classes of antibiotics using whole genome sequence (WGS) data of Pseudomonas aeruginosa.

METHOD

A machine learning technique called Random Forest (RF) and BioWeka was used for classification accuracy assessment and logistic regression (LR) for statistical analysis.

RESULTS

Our results show 44.66% of isolates were resistant to twelve antimicrobial agents and 55.33% were sensitive. The mean classification accuracy was obtained ≥98% for BioWeka and ≥96 for RF on these families of antimicrobials. Where ampicillin was 99.31% and 94.00%, amoxicillin was 99.02% and 95.21%, meropenem was 98.27% and 96.63%, cefepime was 99.73% and 98.34%, fosfomycin was 96.44% and 99.23%, ceftazidime was 98.63% and 94.31%, chloramphenicol was 98.71% and 96.00%, erythromycin was 95.76% and 97.63%, tetracycline was 99.27% and 98.25%, gentamycin was 98.00% and 97.30%, butirosin was 99.57% and 98.03%, and ciprofloxacin was 96.17% and 98.97% with 10-fold-cross validation. In addition, out of twelve, eight drugs have found no false-positive and false-negative bacterial strains.

CONCLUSION

The ability to accurately detect antibiotic resistance could help clinicians make educated decisions about empiric therapy based on the local antibiotic resistance pattern. Moreover, infection prevention may have major consequences if such prescribing practices become widespread for human health.

Lossless indexing with counting de Bruijn graphs

Sequencing data are rapidly accumulating in public repositories. Making this resource accessible for interactive analysis at scale requires efficient approaches for its storage and indexing. There have recently been remarkable advances in building compressed representations of annotated (or colored) de Bruijn graphs for efficiently indexing k-mer sets. However, approaches for representing quantitative attributes such as gene expression or genome positions in a general manner have remained underexplored. In this work, we propose counting de Bruijn graphs, a notion generalizing annotated de Bruijn graphs by supplementing each node-label relation with one or many attributes (e.g., a k-mer count or its positions). Counting de Bruijn graphs index k-mer abundances from 2652 human RNA-seq samples in over eightfold smaller representations compared with state-of-the-art bioinformatics tools and is faster to construct and query. Furthermore, counting de Bruijn graphs with positional annotations losslessly represent entire reads in indexes on average 27% smaller than the input compressed with gzip for human Illumina RNA-seq and 57% smaller for Pacific Biosciences (PacBio) HiFi sequencing of viral samples. A complete searchable index of all viral PacBio SMRT reads from NCBI's Sequence Read Archive (SRA) (152,884 samples, 875 Gbp) comprises only 178 GB. Finally, on the full RefSeq collection, we generate a lossless and fully queryable index that is 4.6-fold smaller than the MegaBLAST index. The techniques proposed in this work naturally complement existing methods and tools using de Bruijn graphs, and significantly broaden their applicability: from indexing k-mer counts and genome positions to implementing novel sequence alignment algorithms on top of highly compressed graph-based sequence indexes.

Explainable Machine Learning for Longitudinal Multi-Omic Microbiome

Over the years, research studies have shown there is a key connection between the microbial community in the gut, genes, and immune system. Understanding this association may help discover the cause of complex chronic idiopathic disorders such as inflammatory bowel disease. Even though important efforts have been put into the field, the functions, dynamics, and causation of dysbiosis state performed by the microbial community remains unclear. Machine learning models can help elucidate important connections and relationships between microbes in the human host. Our study aims to extend the current knowledge of associations between the human microbiome and health and disease through the application of dynamic Bayesian networks to describe the temporal variation of the gut microbiota and dynamic relationships between taxonomic entities and clinical variables. We develop a set of preprocessing steps to clean, filter, select, integrate, and model informative metagenomics, metatranscriptomics, and metabolomics longitudinal data from the Human Microbiome Project. This study accomplishes novel network models with satisfactory predictive performance (accuracy = 0.648) for each inflammatory bowel disease state, validating Bayesian networks as a framework for developing interpretable models to help understand the basic ways the different biological entities (taxa, genes, metabolites) interact with each other in a given environment (human gut) over time. These findings can serve as a starting point to advance the discovery of novel therapeutic approaches and new biomarkers for precision medicine.

SKIOME Project: a curated collection of skin microbiome datasets enriched with study-related metadata

Large amounts of data from microbiome-related studies have been (and are currently being) deposited on international public databases. These datasets represent a valuable resource for the microbiome research community and could serve future researchers interested in integrating multiple datasets into powerful meta-analyses. However, this huge amount of data lacks harmonization and it is far from being completely exploited in its full potential to build a foundation that places microbiome research at the nexus of many subdisciplines within and beyond biology. Thus, it urges the need for data accessibility and reusability, according to findable, accessible, interoperable and reusable (FAIR) principles, as supported by National Microbiome Data Collaborative and FAIR Microbiome. To tackle the challenge of accelerating discovery and advances in skin microbiome research, we collected, integrated and organized existing microbiome data resources from human skin 16S rRNA amplicon-sequencing experiments. We generated a comprehensive collection of datasets, enriched in metadata, and organized this information into data frames ready to be integrated into microbiome research projects and advanced post-processing analyses, such as data science applications (e.g. machine learning). Furthermore, we have created a data retrieval and curation framework built on three different stages to maximize the retrieval of datasets and metadata associated with them. Lastly, we highlighted some caveats regarding metadata retrieval and suggested ways to improve future metadata submissions. Overall, our work resulted in a curated skin microbiome datasets collection accompanied by a state-of-the-art analysis of the last 10 years of the skin microbiome field. Database URL:  https://github.com/giuliaago/SKIOMEMetadataRetrieval.

Extending Association Rule Mining to Microbiome Pattern Analysis: Tools and Guidelines to Support Real Applications

Boosted by the exponential growth of microbiome-based studies, analyzing microbiome patterns is now a hot-topic, finding different fields of application. In particular, the use of machine learning techniques is increasing in microbiome studies, providing deep insights into microbial community composition. In this context, in order to investigate microbial patterns from 16S rRNA metabarcoding data, we explored the effectiveness of Association Rule Mining (ARM) technique, a supervised-machine learning procedure, to extract patterns (in this work, intended as groups of species or taxa) from microbiome data. ARM can generate huge amounts of data, making spurious information removal and visualizing results challenging. Our work sheds light on the strengths and weaknesses of pattern mining strategy into the study of microbial patterns, in particular from 16S rRNA microbiome datasets, applying ARM on real case studies and providing guidelines for future usage. Our results highlighted issues related to the type of input and the use of metadata in microbial pattern extraction, identifying the key steps that must be considered to apply ARM consciously on 16S rRNA microbiome data. To promote the use of ARM and the visualization of microbiome patterns, specifically, we developed microFIM (microbial Frequent Itemset Mining), a versatile Python tool that facilitates the use of ARM integrating common microbiome outputs, such as taxa tables. microFIM implements interest measures to remove spurious information and merges the results of ARM analysis with the common microbiome outputs, providing similar microbiome strategies that help scientists to integrate ARM in microbiome applications. With this work, we aimed at creating a bridge between microbial ecology researchers and ARM technique, making researchers aware about the strength and weaknesses of association rule mining approach.

Elucidating the Beta-Diversity of the Microbiome: from Global Alignment to Local Alignment

Quantitative comparison among microbiomes can link microbial beta-diversity to environmental features, thus enabling prediction of ecosystem properties or dissection of host-microbiome interaction. However, to compute beta-diversity, current methods mainly employ the entire community profiles of taxa or functions, which can miss the subtle differences caused by low-abundance community members that may play crucial roles in the properties of interest. In this work, I review the distance metrics and search engines that we developed to match microbiomes at a large scale based on whole-community-level similarities, as well as their limitations in tackling the microbiome changes caused by less abundant community features. Then I propose the concept of microbiome "local alignment," including an algorithm to measure microbiome similarity on specific fractions of biodiversity and an indexing strategy for rapidly fetching microbiome local-alignment matches from the data repository.

Elucidating the Beta-Diversity of the Microbiome: from Global Alignment to Local Alignment

Quantitative comparison among microbiomes can link microbial beta-diversity to environmental features, thus enabling prediction of ecosystem properties or dissection of host-microbiome interaction. However, to compute beta-diversity, current methods mainly employ the entire community profiles of taxa or functions, which can miss the subtle differences caused by low-abundance community members that may play crucial roles in the properties of interest. In this work, I review the distance metrics and search engines that we developed to match microbiomes at a large scale based on whole-community-level similarities, as well as their limitations in tackling the microbiome changes caused by less abundant community features. Then I propose the concept of microbiome “local alignment,” including an algorithm to measure microbiome similarity on specific fractions of biodiversity and an indexing strategy for rapidly fetching microbiome local-alignment matches from the data repository.

Structure and Function of Oral Microbial Community in Periodontitis Based on Integrated Data

Objective

Microorganisms play a key role in the initiation and progression of periodontal disease. Research studies have focused on seeking specific microorganisms for diagnosing and monitoring the outcome of periodontitis treatment. Large samples may help to discover novel potential biomarkers and capture the common characteristics among different periodontitis patients. This study examines how to screen and merge high-quality periodontitis-related sequence datasets from several similar projects to analyze and mine the potential information comprehensively.

Methods

In all, 943 subgingival samples from nine publications were included based on predetermined screening criteria. A uniform pipeline (QIIME2) was applied to clean the raw sequence datasets and merge them together. Microbial structure, biomarkers, and correlation network were explored between periodontitis and healthy individuals. The microbiota patterns at different periodontal pocket depths were described. Additionally, potential microbial functions and metabolic pathways were predicted using PICRUSt to assess the differences between health and periodontitis.

Results

The subgingival microbial communities and functions in subjects with periodontitis were significantly different from those in healthy subjects. Treponema, TG5, Desulfobulbus, Catonella, Bacteroides, Aggregatibacter, Peptostreptococcus, and Eikenella were periodontitis biomarkers, while Veillonella, Corynebacterium, Neisseria, Rothia, Paludibacter, Capnocytophaga, and Kingella were signature of healthy periodontium. With the variation of pocket depth from shallow to deep pocket, the proportion of Spirochaetes, Bacteroidetes, TM7, and Fusobacteria increased, whereas that of Proteobacteria and Actinobacteria decreased. Synergistic relationships were observed among different pathobionts and negative relationships were noted between periodontal pathobionts and healthy microbiota.

Conclusion

This study shows significant differences in the oral microbial community and potential metabolic pathways between the periodontitis and healthy groups. Our integrated analysis provides potential biomarkers and directions for in-depth research. Moreover, a new method for integrating similar sequence data is shown here that can be applied to other microbial-related areas.

Hierarchical Meta-Storms enables comprehensive and rapid comparison of microbiome functional profiles on a large scale using hierarchical dissimilarity metrics and parallel computing

Functional beta-diversity analysis on numerous microbiomes interprets the linkages between metabolic functions and their meta-data. To evaluate the microbiome beta-diversity, widely used distance metrices only count overlapped gene families but omit their inherent relationships, resulting in erroneous distances due to the sparsity of high-dimensional function profiles. Here we propose Hierarchical Meta-Storms (HMS) to tackle such problem. HMS contains two core components: (i) a dissimilarity algorithm that comprehensively measures functional distances among microbiomes using multi-level metabolic hierarchy and (ii) a fast Principal Co-ordinates Analysis (PCoA) implementation that deduces the beta-diversity pattern optimized by parallel computing. Results showed HMS can detect the variations of microbial functions in upper-level metabolic pathways, however, always missed by other methods. In addition, HMS accomplished the pairwise distance matrix and PCoA for 20 000 microbiomes in 3.9 h on a single computing node, which was 23 times faster and 80% less RAM consumption compared to existing methods, enabling the in-depth data mining among microbiomes on a high resolution. HMS takes microbiome functional profiles as input, produces their pairwise distance matrix and PCoA coordinates.

Availability and implementation

It is coded in C/C++ with parallel computing and released in two alternative forms: a standalone software (https://github.com/qdu-bioinfo/hierarchical-meta-storms) and an equivalent R package (https://github.com/qdu-bioinfo/hrms).

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Towards multi-label classification: Next step of machine learning for microbiome research

Machine learning (ML) has been widely used in microbiome research for biomarker selection and disease prediction. By training microbial profiles of samples from patients and healthy controls, ML classifiers constructs data models by community features that highly correlated with the target diseases, so as to determine the status of new samples. To clearly understand the host-microbe interaction of specific diseases, previous studies always focused on well-designed cohorts, in which each sample was exactly labeled by a single status type. However, in fact an individual may be associated with multiple diseases simultaneously, which introduce additional variations on microbial patterns that interferes the status detection. More importantly, comorbidities or complications can be missed by regular ML models, limiting the practical application of microbiome techniques. In this review, we summarize the typical ML approaches of single-label classification for microbiome research, and demonstrate their limitations in multi-label disease detection using a real dataset. Then we prospect a further step of ML towards multi-label classification that potentially solves the aforementioned problem, including a series of promising strategies and key technical issues for applying multi-label classification in microbiome-based studies.

Holo-omics for deciphering plant-microbiome interactions

Host-microbiome interactions are recognized for their importance to host health. An improved understanding of the molecular underpinnings of host-microbiome relationships will advance our capacity to accurately predict host fitness and manipulate interaction outcomes. Within the plant microbiome research field, unlocking the functional relationships between plants and their microbial partners is the next step to effectively using the microbiome to improve plant fitness. We propose that strategies that pair host and microbial datasets-referred to here as holo-omics-provide a powerful approach for hypothesis development and advancement in this area. We discuss several experimental design considerations and present a case study to highlight the potential for holo-omics to generate a more holistic perspective of molecular networks within the plant microbiome system. In addition, we discuss the biggest challenges for conducting holo-omics studies; specifically, the lack of vetted analytical frameworks, publicly available tools, and required technical expertise to process and integrate heterogeneous data. Finally, we conclude with a perspective on appropriate use-cases for holo-omics studies, the need for downstream validation, and new experimental techniques that hold promise for the plant microbiome research field. We argue that utilizing a holo-omics approach to characterize host-microbiome interactions can provide important opportunities for broadening system-level understandings and significantly inform microbial approaches to improving host health and fitness. Video abstract.

Microbiome Search Engine 2: a Platform for Taxonomic and Functional Search of Global Microbiomes on the Whole-Microbiome Level

A search-based strategy is useful for large-scale mining of microbiome data sets, such as a bird’s-eye view of the microbiome data space and disease diagnosis via microbiome big data. Here, we introduce Microbiome Search Engine 2 (MSE 2), a microbiome database platform for searching query microbiomes against the existing microbiome data sets on the basis of their similarity in taxonomic structure or functional profile.

Metagenomic data sets from diverse environments have been growing rapidly. To ensure accessibility and reusability, tools that quickly and informatively correlate new microbiomes with existing ones are in demand. Here, we introduce Microbiome Search Engine 2 (MSE 2), a microbiome database platform for searching query microbiomes in the global metagenome data space based on the taxonomic or functional similarity of a whole microbiome to those in the database. MSE 2 consists of (i) a well-organized and regularly updated microbiome database that currently contains over 250,000 metagenomic shotgun and 16S rRNA gene amplicon samples associated with unified metadata collected from 798 studies, (ii) an enhanced search engine that enables real-time and fast (<0.5 s per query) searches against the entire database for best-matched microbiomes using overall taxonomic or functional profiles, and (iii) a Web-based graphical user interface for user-friendly searching, data browsing, and tutoring. MSE 2 is freely accessible via http://mse.ac.cn. For standalone searches of customized microbiome databases, the kernel of the MSE 2 search engine is provided at GitHub (https://github.com/qibebt-bioinfo/meta-storms).

IMPORTANCE A search-based strategy is useful for large-scale mining of microbiome data sets, such as a bird’s-eye view of the microbiome data space and disease diagnosis via microbiome big data. Here, we introduce Microbiome Search Engine 2 (MSE 2), a microbiome database platform for searching query microbiomes against the existing microbiome data sets on the basis of their similarity in taxonomic structure or functional profile. Key improvements include database extension, data compatibility, a search engine kernel, and a user interface. The new ability to search the microbiome space via functional similarity greatly expands the scope of search-based mining of the microbiome big data.

Genomes OnLine Database (GOLD) v.8: overview and updates

The Genomes OnLine Database (GOLD) (https://gold.jgi.doe.gov/) is a manually curated, daily updated collection of genome projects and their metadata accumulated from around the world. The current version of the database includes over 1.17 million entries organized broadly into Studies (45 770), Organisms (387 382) or Biosamples (101 207), Sequencing Projects (355 364) and Analysis Projects (283 481). These four levels contain over 600 metadata fields, which includes 76 controlled vocabulary (CV) tables containing 3873 terms. GOLD provides an interactive web user interface for browsing and searching by a wide range of project and metadata fields. Users can enter details about their own projects in GOLD, which acts as a gatekeeper to ensure that metadata is accurately documented before submitting sequence information to the Integrated Microbial Genomes (IMG) system for analysis. In order to maintain a reference dataset for use by members of the scientific community, GOLD also imports projects from public repositories such as GenBank and SRA. The current status of the database, along with recent updates and improvements are described in this manuscript.

Meta-Apo improves accuracy of 16S-amplicon-based prediction of microbiome function

BACKGROUND

Due to their much lower costs in experiment and computation than metagenomic whole-genome sequencing (WGS), 16S rRNA gene amplicons have been widely used for predicting the functional profiles of microbiome, via software tools such as PICRUSt 2. However, due to the potential PCR bias and gene profile variation among phylogenetically related genomes, functional profiles predicted from 16S amplicons may deviate from WGS-derived ones, resulting in misleading results.

RESULTS

Here we present Meta-Apo, which greatly reduces or even eliminates such deviation, thus deduces much more consistent diversity patterns between the two approaches. Tests of Meta-Apo on > 5000 16S-rRNA amplicon human microbiome samples from 4 body sites showed the deviation between the two strategies is significantly reduced by using only 15 WGS-amplicon training sample pairs. Moreover, Meta-Apo enables cross-platform functional comparison between WGS and amplicon samples, thus greatly improve 16S-based microbiome diagnosis, e.g. accuracy of gingivitis diagnosis via 16S-derived functional profiles was elevated from 65 to 95% by WGS-based classification. Therefore, with the low cost of 16S-amplicon sequencing, Meta-Apo can produce a reliable, high-resolution view of microbiome function equivalent to that offered by shotgun WGS.

CONCLUSIONS

This suggests that large-scale, function-oriented microbiome sequencing projects can probably benefit from the lower cost of 16S-amplicon strategy, without sacrificing the precision in functional reconstruction that otherwise requires WGS. An optimized C++ implementation of Meta-Apo is available on GitHub ( https://github.com/qibebt-bioinfo/meta-apo ) under a GNU GPL license. It takes the functional profiles of a few paired WGS:16S-amplicon samples as training, and outputs the calibrated functional profiles for the much larger number of 16S-amplicon samples.