Metagenomic Datasets Research Articles

MAGqual: a stand-alone pipeline to assess the quality of metagenome-assembled genomes

BackgroundMetagenomics, the whole genome sequencing of microbial communities, has provided insight into complex ecosystems. It has facilitated the discovery of novel microorganisms, explained community interactions and found applications in various fields. Advances in high-throughput and third-generation sequencing technologies have further fuelled its popularity. Nevertheless, managing the vast data produced and addressing variable dataset quality remain ongoing challenges. Another challenge arises from the number of assembly and binning strategies used across studies. Comparing datasets and analysis tools is complex as it requires the quantitative assessment of metagenome quality. The inherent limitations of metagenomic sequencing, which often involves sequencing complex communities, mean community members are challenging to interrogate with traditional culturing methods leading to many lacking reference sequences. MIMAG standards aim to provide a method to assess metagenome quality for comparison but have not been widely adopted.ResultsTo address the need for simple and quick metagenome quality assignation, here we introduce the pipeline MAGqual (Metagenome-Assembled Genome qualifier) and demonstrate its effectiveness at determining metagenomic dataset quality in the context of the MIMAG standards.ConclusionsThe MAGqual pipeline offers an accessible way to evaluate metagenome quality and generate metadata on a large scale. MAGqual is built in Snakemake to ensure readability and scalability, and its open-source nature promotes accessibility, community development, and ease of updates. MAGqual is built in Snakemake, R, and Python and is available under the MIT license on GitHub at https://github.com/ac1513/MAGqual.EB_9RbKd3CYdEq3zCTHKaqVideo

Variation in the metagenomic analysis of fecal microbiome composition calls for a standardized operating approach.

The reproducibility in microbiome studies is limited due to the lack of one gold-standard operating procedure. The aim of this study was to examine the impact of protocol variations on microbiome composition using metagenomic data sets from a single center. We assessed the variation in a data set consisted of 2,722 subjects, including 9 subcohorts harboring healthy subjects and patients with various disorders, such as inflammatory bowel disease, colorectal cancer, and type 2 diabetes. Two different DNA extraction kits, with or without lyticase, and two sample storage methods were compared. Our results indicated that DNA extraction had the largest impact on gut microbiota diversity among all host factors and sample operating procedures. Healthy subjects matched by age, body mass index, and sample operating methods exhibited reduced, yet significant differences (PERMANOVA, P < 0.05) in gut microbiota composition across studies. The variations contributed by DNA extraction were primarily driven by different recovery efficiency of gram-positive bacteria, e.g., phyla Firmicutes and Actinobacteria. This was further confirmed by a parallel comparison of fecal samples from five healthy subjects and a standard mock community. In addition, the DNA extraction method influenced DNA biomass, quality, and the detection of specific lineage-associated diseases. Sample operating approach and batch effects should be considered for cohorts with large sample size or longitudinal cohorts to ensure that source data were appropriately generated and analyzed. Comparison between samples processed with inconsistent methods should be dealt with caution. This study will promote the establishment of a sample operating standard to enhance our understanding of microbiome and translating in clinical practice.IMPORTANCEThe reproducibility of human gut microbiome studies has been suboptimal across cohorts and study design choices. One possible reason for the disagreement is the introduction of systemic biases due to differences in methodologies. In our study, we utilized microbial metagenomic data sets from 2,722 fecal samples generated from a single research center to examine the extent to which sample storage and DNA extraction influence the quantification of microbial composition and compared this variable with other sources of technical and biological variation. Our research highlights the impact of DNA extraction methods when analyzing microbiome data and suggests that the microbiome profile may be influenced by differences in the extraction efficiency of bacterial species. With metagenomics sequencing being increasingly used in clinical biology, our findings provide insight into the challenges using metagenomics sequencing in clinical diagnostics, where the detection of certain species and its abundance relative to a "healthy reference" is key.

Rapid identification of antibiotic resistance gene hosts by prescreening ARG-like reads

Effective risk assessment and control of environmental antibiotic resistance depend on comprehensive information about antibiotic resistance genes (ARGs) and their microbial hosts. Advances in sequencing technologies and bioinformatics have enabled the identification of ARG hosts using metagenome-assembled contigs and genomes. However, these approaches often suffer from information loss and require extensive computational resources. Here we introduce a bioinformatic strategy that identifies ARG hosts by prescreening ARG-like reads (ALRs) directly from total metagenomic datasets. This ALR-based method offers several advantages: (1) it enables the detection of low-abundance ARG hosts with higher accuracy in complex environments; (2) it establishes a direct relationship between the abundance of ARGs and their hosts; and (3) it reduces computation time by approximately 44–96% compared to strategies relying on assembled contigs and genomes. We applied our ALR-based strategy alongside two traditional methods to investigate a typical human-impacted environment. The results were consistent across all methods, revealing that ARGs are predominantly carried by Gammaproteobacteria and Bacilli, and their distribution patterns may indicate the impact of wastewater discharge on coastal resistome. Our strategy provides rapid and accurate identification of antibiotic-resistant bacteria, offering valuable insights for the high-throughput surveillance of environmental antibiotic resistance. This study further expands our knowledge of ARG-related risk management in future.

Aquificae overcomes competition by archaeal thermophiles, and crowding by bacterial mesophiles, to dominate the boiling vent-water of a Trans-Himalayan sulfur-borax spring.

Trans-Himalayan hot spring waters rich in boron, chlorine, sodium and sulfur (but poor in calcium and silicon) are known based on PCR-amplified 16S rRNA gene sequence data to harbor high diversities of infiltrating bacterial mesophiles. Yet, little is known about the community structure and functions, primary productivity, mutual interactions, and thermal adaptations of the microorganisms present in the steaming waters discharged by these geochemically peculiar spring systems. We revealed these aspects of a bacteria-dominated microbiome (microbial cell density ~8.5 × 104 mL-1; live:dead cell ratio 1.7) thriving in the boiling (85°C) fluid vented by a sulfur-borax spring called Lotus Pond, situated at 4436 m above the mean sea-level, in the Puga valley of eastern Ladakh, on the Changthang plateau. Assembly, annotation, and population-binning of >15-GB metagenomic sequence illuminated the numeral predominance of Aquificae. While members of this phylum accounted for 80% of all 16S rRNA-encoding reads within the metagenomic dataset, 14% of such reads were attributed to Proteobacteria. Post assembly, only 25% of all protein-coding genes identified were attributable to Aquificae, whereas 41% was ascribed to Proteobacteria. Annotation of metagenomic reads encoding 16S rRNAs, and/or PCR-amplified 16S rRNA genes, identified 163 bacterial genera, out of which 66 had been detected in past investigations of Lotus Pond's vent-water via 16S amplicon sequencing. Among these 66, Fervidobacterium, Halomonas, Hydrogenobacter, Paracoccus, Sulfurihydrogenibium, Tepidimonas, Thermus and Thiofaba (or their close phylogenomic relatives) were presently detected as metagenome-assembled genomes (MAGs). Remarkably, the Hydrogenobacter related MAG alone accounted for ~56% of the entire metagenome, even though only 15 out of the 66 genera consistently present in Lotus Pond's vent-water have strains growing in the laboratory at >45°C, reflecting the continued existence of the mesophiles in the ecosystem. Furthermore, the metagenome was replete with genes crucial for thermal adaptation in the context of Lotus Pond's geochemistry and topography. In terms of sequence similarity, a majority of those genes were attributable to phylogenetic relatives of mesophilic bacteria, while functionally they rendered functions such as encoding heat shock proteins, molecular chaperones, and chaperonin complexes; proteins controlling/modulating/inhibiting DNA gyrase; universal stress proteins; methionine sulfoxide reductases; fatty acid desaturases; different toxin-antitoxin systems; enzymes protecting against oxidative damage; proteins conferring flagellar structure/function, chemotaxis, cell adhesion/aggregation, biofilm formation, and quorum sensing. The Lotus Pond Aquificae not only dominated the microbiome numerically but also acted potentially as the main primary producers of the ecosystem, with chemolithotrophic sulfur oxidation (Sox) being the fundamental bioenergetic mechanism, and reductive tricarboxylic acid (rTCA) cycle the predominant carbon fixation pathway. The Lotus Pond metagenome contained several genes directly or indirectly related to virulence functions, biosynthesis of secondary metabolites including antibiotics, antibiotic resistance, and multi-drug efflux pumping. A large proportion of these genes being attributable to Aquificae, and Proteobacteria (very few were ascribed to Archaea), it could be worth exploring in the future whether antibiosis helped the Aquificae overcome niche overlap with other thermophiles (especially those belonging to Archaea), besides exacerbating the bioenergetic costs of thermal endurance for the mesophilic intruders of the ecosystem.

Assessing antimicrobial resistance in pasture-based dairy farms: a 15-month surveillance study in New Zealand.

Antimicrobial resistance is a global public and animal health concern. Antimicrobial resistance genes (ARGs) have been detected in dairy farm environments globally; however, few longitudinal studies have utilized shotgun metagenomics for ARG surveillance in pasture-based systems. This 15-month study aimed to undertake a baseline survey using shotgun metagenomics to assess the relative abundance and diversity of ARGs in two pasture-based dairy farm environments in New Zealand with different management practices. There was no statistically significant difference in overall ARG relative abundance between the two dairy farms (P = 0.321) during the study period. Compared with overseas data, the relative abundance of ARG copies per 16S rRNA gene in feces (0.08-0.17), effluent (0.03-0.37), soil (0.20-0.63), and bulk tank milk (0.0-0.12) samples was low. Models comparing the presence or absence of resistance classes found in >10% of all feces, effluent, and soil samples demonstrated no statistically significant associations (P > 0.05) with "season," and only multi-metal (P = 0.020) and tetracycline (P = 0.0003) resistance were significant at the "farm" level. Effluent samples harbored the most diverse ARGs, some with a recognized public health risk, whereas soil samples had the highest ARG relative abundance but without recognized health risks. This highlights the importance of considering the genomic context and risk of ARGs in metagenomic data sets. This study suggests that antimicrobial resistance on pasture-based dairy farms is low and provides essential baseline ARG surveillance data for such farming systems.IMPORTANCEAntimicrobial resistance is a global threat to human and animal health. Despite the detection of antimicrobial resistance genes (ARGs) in dairy farm environments globally, longitudinal surveillance in pasture-based systems remains limited. This study assessed the relative abundance and diversity of ARGs in two New Zealand dairy farms with different management practices and provided important baseline ARG surveillance data on pasture-based dairy farms. The overall ARG relative abundance on these two farms was low, which provides further evidence for consumers of the safety of New Zealand's export products. Effluent samples harbored the most diverse range of ARGs, some of which were classified with a recognized risk to public health, whereas soil samples had the highest ARG relative abundance; however, the soil ARGs were not classified with a recognized public health risk. This emphasizes the need to consider genomic context and risk as well as ARG relative abundance in resistome studies.

Species- and subspecies-level characterization of health-associated bacterial consortia that colonize the human gut during infancy

ABSTRACT Background The human gut microbiome develops rapidly during infancy, a key window of development coinciding with the maturation of the adaptive immune system. However, little is known about the microbiome growth dynamics over the first few months of life and whether there are any generalizable patterns across human populations. We performed metagenomic sequencing on stool samples (n = 94) from a cohort of infants (n = 15) at monthly intervals in the first 6 months of life, augmenting our dataset with seven published studies for a total of 4,441 metagenomes from 1,162 infants. Results Strain-level de novo analysis was used to identify 592 of the most abundant organisms in the infant gut microbiome. Previously unrecognized consortia were identified which exhibited highly correlated abundances across samples and were composed of diverse species spanning multiple genera. Analysis of a published cohort of infants with cystic fibrosis identified one such novel consortium of diverse Enterobacterales which was positively correlated with weight gain. While all studies showed an increased community stability during the first year of life, microbial dynamics varied widely in the first few months of life, both by study and by individual. Conclusion By augmenting published metagenomic datasets with data from a newly established cohort, we were able to identify novel groups of organisms that are correlated with measures of robust human development. We hypothesize that the presence of these groups may impact human health in aggregate in ways that individual species may not in isolation.

DeepPL: A deep-learning-based tool for the prediction of bacteriophage lifecycle.

Bacteriophages (phages) are viruses that infect bacteria and can be classified into two different lifecycles. Virulent phages (or lytic phages) have a lytic cycle that can lyse the bacteria host after their infection. Temperate phages (or lysogenic phages) can integrate their phage genomes into bacterial chromosomes and replicate with bacterial hosts via the lysogenic cycle. Identifying phage lifecycles is a crucial step in developing suitable applications for phages. Compared to the complicated traditional biological experiments, several tools have been designed for predicting phage lifecycle using different algorithms, such as random forest (RF), linear support-vector classifier (SVC), and convolutional neural network (CNN). In this study, we developed a natural language processing (NLP)-based tool-DeepPL-for predicting phage lifecycles via nucleotide sequences. The test results showed that our DeepPL had an accuracy of 94.65% with a sensitivity of 92.24% and a specificity of 95.91%. Moreover, DeepPL had 100% accuracy in lifecycle prediction on the phages we isolated and biologically verified previously in the lab. Additionally, a mock phage community metagenomic dataset was used to test the potential usage of DeepPL in viral metagenomic research. DeepPL displayed a 100% accuracy for individual phage complete genomes and high accuracies ranging from 71.14% to 100% on phage contigs produced by various next-generation sequencing technologies. Overall, our study indicates that DeepPL has a reliable performance on phage lifecycle prediction using the most fundamental nucleotide sequences and can be applied to future phage and metagenomic research.

Identification of Shemin pathway genes for tetrapyrrole biosynthesis in bacteriophage sequences from aquatic environments

Tetrapyrroles such as heme, chlorophyll, and vitamin B12 are essential for various metabolic pathways. They derive from 5-aminolevulinic acid (5-ALA), which can be synthesized by a single enzyme (5-ALA synthase or AlaS, Shemin pathway) or by a two-enzyme pathway. The genomes of some bacteriophages from aquatic environments carry various tetrapyrrole biosynthesis genes. Here, we analyze available metagenomic datasets and identify alaS homologs (viral alaS, or valaS) in sequences corresponding to marine and freshwater phages. The genes are found individually or as part of complete or truncated three-gene loci encoding heme-catabolizing enzymes. Amino-acid sequence alignments and three-dimensional structure prediction support that the valaS sequences likely encode functional enzymes. Indeed, we demonstrate that is the case for a freshwater phage valaS sequence, as it can complement an Escherichia coli 5-ALA auxotroph, and an E. coli strain overexpressing the gene converts the typical AlaS substrates glycine and succinyl-CoA into 5-ALA. Thus, our work identifies valaS as an auxiliary metabolic gene in phage sequences from aquatic environments, further supporting the importance of tetrapyrrole metabolism in bacteriophage biology.

Complete genomes of Asgard archaea reveal diverse integrated and mobile genetic elements.

Asgard archaea are of great interest as the progenitors of Eukaryotes, but little is known about the mobile genetic elements (MGEs) that may shape their ongoing evolution. Here, we describe MGEs that replicate in Atabeyarchaeia, a wetland Asgard archaea lineage represented by two complete genomes. We used soil depth-resolved population metagenomic data sets to track 18 MGEs for which genome structures were defined and precise chromosome integration sites could be identified for confident host linkage. Additionally, we identified a complete 20.67 kbp circular plasmid and two family-level groups of viruses linked to Atabeyarchaeia, via CRISPR spacer targeting. Closely related 40 kbp viruses possess a hypervariable genomic region encoding combinations of specific genes for small cysteine-rich proteins structurally similar to restriction-homing endonucleases. One 10.9 kbp integrative conjugative element (ICE) integrates genomically into the Atabeyarchaeum deiterrae-1 chromosome and has a 2.5 kbp circularizable element integrated within it. The 10.9 kbp ICE encodes an expressed Type IIG restriction-modification system with a sequence specificity matching an active methylation motif identified by Pacific Biosciences (PacBio) high-accuracy long-read (HiFi) metagenomic sequencing. Restriction-modification of Atabeyarchaeia differs from that of another coexisting Asgard archaea, Freyarchaeia, which has few identified MGEs but possesses diverse defense mechanisms, including DISARM and Hachiman, not found in Atabeyarchaeia. Overall, defense systems and methylation mechanisms of Asgard archaea likely modulate their interactions with MGEs, and integration/excision and copy number variation of MGEs in turn enable host genetic versatility.

Graph-based self-supervised learning for repeat detection in metagenomic assembly.

Repetitive DNA (repeats) poses significant challenges for accurate and efficient genome assembly and sequence alignment. This is particularly true for metagenomic data, in which genome dynamics such as horizontal gene transfer, gene duplication, and gene loss/gain complicate accurate genome assembly from metagenomic communities. Detecting repeats is a crucial first step in overcoming these challenges. To address this issue, we propose GraSSRep, a novel approach that leverages the assembly graph's structure through graph neural networks (GNNs) within a self-supervised learning framework to classify DNA sequences into repetitive and nonrepetitive categories. Specifically, we frame this problem as a node classification task within a metagenomic assembly graph. In a self-supervised fashion, we rely on a high-precision (but low-recall) heuristic to generate pseudolabels for a small proportion of the nodes. We then use those pseudolabels to train a GNN embedding and a random forest classifier to propagate the labels to the remaining nodes. In this way, GraSSRep combines sequencing features with predefined and learned graph features to achieve state-of-the-art performance in repeat detection. We evaluate our method using simulated and synthetic metagenomic data sets. The results on the simulated data highlight GraSSRep's robustness to repeat attributes, demonstrating its effectiveness in handling the complexity of repeated sequences. Additionally, experiments with synthetic metagenomic data sets reveal that incorporating the graph structure and the GNN enhances the detection performance. Finally, in comparative analyses, GraSSRep outperforms existing repeat detection tools with respect to precision and recall.

Monitoring and preventing foodborne outbreaks: are we missing wastewater as a key data source?

In 2022, the number of foodborne outbreaks in Europe increased by 43.9%, highlighting the need to improve surveillance systems and design outbreak predictive tools. This review aims to assess the scientific literature describing wastewater surveillance to monitor foodborne pathogens in association with clinical data. In the selected studies, the relationship between peaks of pathogen concentration in wastewater and reported clinical cases is described. Moreover, details on analytical methods to detect and quantify pathogens as well as wastewater sampling procedures are discussed. Few papers show a statistically significant correlation between high concentrations of foodborne pathogens in wastewater and the occurrence of clinical cases. However, monitoring pathogen concentration in wastewater looks like a promising and cost-effective strategy to improve foodborne outbreak surveillance. Such a strategy can be articulated in three steps, where the first one is testing wastewater with an untargeted method, like shotgun metagenomic, to detect microorganisms belonging to different domains. The second consists of testing wastewater with a targeted method, such as quantitative polymerase chain reaction, to quantify those specific pathogens that in the metagenomic dataset display an increasing trend or exceed baseline concentration thresholds. The third involves the integrated wastewater and clinical data analysis and modeling to find meaningful epidemiological correlations and make predictions.

Unveiling the mycobiome of ecological importance from termitarium soils

BackgroundTermitarium is an intricate biological home with exceptional physicochemical soil properties and numerous beneficial microbiomes. The termitarium microbiome plays a role in several bionetwork processes. It contributes significantly to sustainable agriculture, forestry, and horticulture. In contrast to the immense number of research works focusing on termitarium bacteria and archaea, the amount of study about termitarium fungal communities is limited. This is despite the fact that fungi significantly contribute to the cycling of energy and matter on Earth. Thus, this study investigates the metagenomic datasets from termitarium soil collected from agricultural soils, to obtain a depiction of their mycobiome of ecological importance.ResultsOur dataset showed that Ascomycota was the most rank phylum, while Basidiomycota was the next in rank. Specifically, we identified numerous fungi of commercially importance, for example, producers of cellulases and cellulosomes. This was strengthened with the presence of functional genes/enzymes capable of promoting plant growth traits and defence system. Analysis further shown that soil properties influenced the distribution of fungal communities. Magnesium, nitrogen, and potassium, for example, were observed to be among the best predictor of fungal distribution and explained 87.40% of the total variation.ConclusionThis study epitomize further insights into the mycobiome of termitarium soils that are of ecological interest. Again, with some of our sequences assembled as uncategorized fungi, further studies are recommended for the evaluation of the metabolic potentials of these fungi as there is a possibility of the discovery of new fungi or genes of industrial or environmental applications.

Global meta‐analysis reveals the drivers of gut microbiome variation across vertebrates

AbstractShifts in gut microbial diversity and structure are one route by which vertebrate hosts adapt to local environmental conditions. However, recent studies have mostly been limited to a single species, small sample sizes, or restricted geographic ranges. Therefore, drawing a global picture of vertebrate gut microbiome diversity, community structure, and determinants for their adaptive shifts remains to be elucidated. We here collected 6508 samples from 113 vertebrate species covering diverse classes, feeding behaviors, and host habitats based on 16S rRNA gene sequencing. The results showed that host diet pattern had a significant impact on gut microbiome variation, which might drive taxonomic and functional contents of gut microbiome across vertebrates. Of note, the phylum Fusobacteria were enriched in carnivorous vertebrate gut while herbivorous vertebrate gut selectively increased the abundance of Verrucomicrobia. Also, climate factors were strongly associated with gut microbiome variation across vertebrates. Interestingly, we found that the abundance of microbiota belonging to Bacteroidetes increased gradually while the members from Proteobacteria showed a decreasing trend from high‐ to low‐latitude zones, potentially contributing to vertebrate adaptation to local climate condition. Additionally, we comprehensively deciphered the common antibiotic resistomes and their potential mobility between terrestrial vertebrate gut microbiome (n = 487) and their sympatric soil biological environment samples (n = 203) by integrating metagenomic sequencing datasets. Particularly, potential horizontal antibiotic resistance genes (e.g., bacA) transfers were detected between vertebrates gut microbiome and their sympatric soil biological environment. Together, our findings provide new evidence of how external environmental factors affect vertebrate gut microbiome variation.

Understanding the mechanism of microplastic-associated antibiotic resistance genes in aquatic ecosystems: Insights from metagenomic analyses and machine learning

The pervasive presence of microplastics (MPs) in aquatic systems facilitates the transmission of antibiotic resistance genes (ARGs), thereby posing risks to ecosystems and human well-being. However, owing to variations in environmental backgrounds and the limited scope of research subjects, studies on ARGs in MPs lack unified conclusions, particularly regarding whether different types of MPs selectively promote ARG enrichment. Analysing large-scale datasets can better encompass broad spatiotemporal scales and diverse samples, facilitating a more extensive exploration of the complex ecological relationships between MPs and ARGs. The present study integrated existing metagenomic datasets to conduct a comprehensive risk assessment and comparative analysis of resistance groups across various MPs. In addition, we endeavoured to elucidate potential associations between ARGs and bacterial taxa, as well as MP structural features, using machine learning (ML) methods. The findings of our research highlight the pivotal role of MP type in shaping plastispheres, accounting for 9.56 % of the biotic variation (Adonis index) and explaining 18.59 % of the ARG variance. Compared to conventional MPs, biodegradable MPs, such as polyhydroxyalkanoates (PHA) and polylactic acid (PLA), exhibit lower species uniformity and diversity but pose a higher risk of ARG occurrence. These ML approaches effectively forecasted ARG abundance by using the bacterial taxa and molecular structure descriptors (MDs) of MPs (average R2tra = 0.882, R2test = 0.759). Feature analysis showed that MDs associated with lipophilicity, solubility, toxicity, and surface potential significantly influenced the relative abundance of ARGs in the plastispheres. The interpretable multiple linear regression (MLR) model, particularly notable, elucidated a linear relationship between bacterial genera and ARGs, offering promise for identifying potential ARG hosts. This study offers novel insights into ARG dynamics and ecological risks within aquatic plastispheres, highlighting the importance of comprehensive MP monitoring initiatives.

A core microbiome signature as an indicator of health

The gut microbiota is crucial for human health, functioning as a complex adaptive system akin to a vital organ. To identify core health-relevant gut microbes, we followed the systems biology tenet that stable relationships signify core components. By analyzing metagenomic datasets from a high-fiber dietary intervention in type 2 diabetes and 26 case-control studies across 15 diseases, we identified a set of stably correlated genome pairs within co-abundance networks perturbed by dietary interventions and diseases. These genomes formed a "two competing guilds" (TCGs) model, with one guild specialized in fiber fermentation and butyrate production and the other characterized by virulence and antibiotic resistance. Our random forest models successfully distinguished cases from controls across multiple diseases and predicted immunotherapy outcomes through the use of these genomes. Our guild-based approach, which is genome specific, database independent, and interaction focused, identifies a core microbiome signature that serves as a holistic health indicator and a potential common target for health enhancement.

MNBC: a multithreaded Minimizer-based Naïve Bayes Classifier for improved metagenomic sequence classification.

State-of-the-art tools for classifying metagenomic sequencing reads provide both rapid and accurate options, although the combination of both in a single tool is a constantly improving area of research. The machine learning-based Naïve Bayes Classifier (NBC) approach provides a theoretical basis for accurate classification of all reads in a sample. We developed the multithreaded Minimizer-based Naïve Bayes Classifier (MNBC) tool to improve the NBC approach by applying minimizers, as well as plurality voting for closely related classification scores. A standard reference- and test-sequence framework using simulated variable-length reads benchmarked MNBC with six other state-of-the-art tools: MetaMaps, Ganon, Kraken2, KrakenUniq, CLARK, and Centrifuge. We also applied MNBC to the "marine" and "strain-madness" short-read metagenomic datasets in the Critical Assessment of Metagenome Interpretation (CAMI) II challenge using a corresponding database from the time. MNBC efficiently identified reads from unknown microorganisms, and exhibited the highest species- and genus-level precision and recall on short reads, as well as the highest species-level precision on long reads. It also achieved the highest accuracy on the "strain-madness" dataset. MNBC is freely available at: https://github.com/ComputationalPathogens/MNBC.

Strain-resolved de-novo metagenomic assembly of viral genomes and microbial 16S rRNAs

BackgroundMetagenomics is a powerful approach to study environmental and human-associated microbial communities and, in particular, the role of viruses in shaping them. Viral genomes are challenging to assemble from metagenomic samples due to their genomic diversity caused by high mutation rates. In the standard de Bruijn graph assemblers, this genomic diversity leads to complex k-mer assembly graphs with a plethora of loops and bulges that are challenging to resolve into strains or haplotypes because variants more than the k-mer size apart cannot be phased. In contrast, overlap assemblers can phase variants as long as they are covered by a single read.ResultsHere, we present PenguiN, a software for strain resolved assembly of viral DNA and RNA genomes and bacterial 16S rRNA from shotgun metagenomics. Its exhaustive detection of all read overlaps in linear time combined with a Bayesian model to select strain-resolved extensions allow it to assemble longer viral contigs, less fragmented genomes, and more strains than existing assembly tools, on both real and simulated datasets. We show a 3–40-fold increase in complete viral genomes and a 6-fold increase in bacterial 16S rRNA genes.ConclusionPenguiN is the first overlap-based assembler for viral genome and 16S rRNA assembly from large and complex metagenomic datasets, which we hope will facilitate studying the key roles of viruses in microbial communities.C22dm7AjjRZTtiuc33W1snVideo

Gut metagenome-derived image augmentation and deep learning improve prediction accuracy of metabolic disease classification.

In recent years, statistics and machine learning methods have been widely used to analyze the relationship between human gut microbial metagenome and metabolic diseases, which is of great significance for the functional annotation and development of microbial communities. In this study, we proposed a new and scalable framework for image enhancement and deep learning of gut metagenome, which could be used in the classification of human metabolic diseases. Each data sample in three representative human gut metagenome datasets was transformed into image and enhanced, and put into the machine learning models of logistic regression (LR), support vector machine (SVM), Bayesian network (BN) and random forest (RF), and the deep learning models of multilayer perceptron (MLP) and convolutional neural network (CNN). The accuracy performance of the overall evaluation model for disease prediction was verified by accuracy (A), accuracy (P), recall (R), F1 score (F1), area under ROC curve (AUC) and 10 fold cross-validation. The results showed that the overall performance of MLP model was better than that of CNN, LR, SVM, BN, RF and PopPhy-CNN, and the performance of MLP and CNN models was further improved after data enhancement (random rotation and adding salt-and-pepper noise). The accuracy of MLP model in disease prediction was further improved by 4%-11%, F1 by 1%-6% and AUC by 5%-10%. The above results showed that human gut metagenome image enhancement and deep learning could accurately extract microbial characteristics and effectively predict the host disease phenotype. The source code and datasets used in this study can be publicly accessed in https://github.com/HuaXWu/GM_ML_Classification.git.

Taxometer: Improving taxonomic classification of metagenomics contigs

For taxonomy based classification of metagenomics assembled contigs, current methods use sequence similarity to identify their most likely taxonomy. However, in the related field of metagenomic binning, contigs are routinely clustered using information from both the contig sequences and their abundance. We introduce Taxometer, a neural network based method that improves the annotations and estimates the quality of any taxonomic classifier using contig abundance profiles and tetra-nucleotide frequencies. We apply Taxometer to five short-read CAMI2 datasets and find that it increases the average share of correct species-level contig annotations of the MMSeqs2 tool from 66.6% to 86.2%. Additionally, it reduce the share of wrong species-level annotations in the CAMI2 Rhizosphere dataset by an average of two-fold for Metabuli, Centrifuge, and Kraken2. Futhermore, we use Taxometer for benchmarking taxonomic classifiers on two complex long-read metagenomics data sets where ground truth is not known. Taxometer is available as open-source software and can enhance any taxonomic annotation of metagenomic contigs.

A realistic benchmark for differential abundance testing and confounder adjustment in human microbiome studies

BackgroundIn microbiome disease association studies, it is a fundamental task to test which microbes differ in their abundance between groups. Yet, consensus on suitable or optimal statistical methods for differential abundance testing is lacking, and it remains unexplored how these cope with confounding. Previous differential abundance benchmarks relying on simulated datasets did not quantitatively evaluate the similarity to real data, which undermines their recommendations.ResultsOur simulation framework implants calibrated signals into real taxonomic profiles, including signals mimicking confounders. Using several whole meta-genome and 16S rRNA gene amplicon datasets, we validate that our simulated data resembles real data from disease association studies much more than in previous benchmarks. With extensively parametrized simulations, we benchmark the performance of nineteen differential abundance methods and further evaluate the best ones on confounded simulations. Only classic statistical methods (linear models, the Wilcoxon test, t-test), limma, and fastANCOM properly control false discoveries at relatively high sensitivity. When additionally considering confounders, these issues are exacerbated, but we find that adjusted differential abundance testing can effectively mitigate them. In a large cardiometabolic disease dataset, we showcase that failure to account for covariates such as medication causes spurious association in real-world applications.ConclusionsTight error control is critical for microbiome association studies. The unsatisfactory performance of many differential abundance methods and the persistent danger of unchecked confounding suggest these contribute to a lack of reproducibility among such studies. We have open-sourced our simulation and benchmarking software to foster a much-needed consolidation of statistical methodology for microbiome research.