Systematic evaluation of supervised machine learning for sample origin prediction using metagenomic sequencing data

Artificial Intelligence: A Promising Tool in Exploring the Phytomicrobiome in Managing Disease and Promoting Plant Health

There is increasing interest in harnessing the microbiome to improve cropping systems. With the availability of high-throughput and low-cost sequencing technologies, gathering microbiome data is becoming more routine. However, the analysis of microbiome data is challenged by the size and complexity of the data, and the incomplete nature of many microbiome databases. Further, to bring microbiome data value, it often needs to be analyzed in conjunction with other complex data that impact on crop health and disease management, such as plant genotype and environmental factors. Artificial intelligence (AI), boosted through deep learning (DL), has achieved significant breakthroughs and is a powerful tool for managing large complex datasets such as the interplay between the microbiome, crop plants, and their environment. In this review, we aim to provide readers with a brief introduction to AI techniques, and we introduce how AI has been applied to areas of microbiome sequencing taxonomy, the functional annotation for microbiome sequences, associating the microbiome community with host traits, designing synthetic communities, genomic selection, field phenotyping, and disease forecasting. At the end of this review, we proposed further efforts that are required to fully exploit the power of AI in studying phytomicrobiomes.

Nitrogen-metabolising microorganism analysis in rapid sand filters from drinking water treatment plant

Sand filters (SFs) are common treatment processes for nitrogen pollutant removal in drinking water treatment plants (DWTPs). However, the mechanisms on the nitrogen-cycling role of SFs are still unclear. In this study, 16S rRNA gene amplicon sequencing was used to characterise the diversity and composition of the bacterial community in SFs from DWTPs. Additionally, metagenomics approach was used to determine the functional microorganisms involved in nitrogen cycle in SFs. Our results showed that Pseudomonadota, Acidobacteria, Nitrospirae and Chloroflexi dominated in SFs. Subsequently, 85 high-quality metagenome-assembled genomes (MAGs) were retrieved from metagenome datasets of selected SFs involving nitrification, assimilatory nitrogen reduction, denitrification and anaerobic ammonia oxidation (anammox) processes. Read mapping to reference genomes of Nitrospira and the phylogenetic tree of the ammonia monooxygenase subunit A gene, amoA, suggested that Nitrospira is abundantly found in SFs. Furthermore, according to their genetic content, a nitrogen metabolic model in SFs was proposed using representative MAGs and pure culture isolate. Quantitative real-time polymerase chain reaction (qPCR) showed that ammonia-oxidising bacteria (AOB) and archaea (AOA), and complete ammonia oxidisers (comammox) were ubiquitous in the SFs, with the abundance of comammox being higher than that of AOA and AOB. Moreover, we identified a bacterial strain with a high NO₃-N removal rate as Pseudomonas sp. DW-5, which could be applied in the bioremediation of micro-polluted drinking water sources. Our study provides insights into functional nitrogen-metabolising microbes in SFs of DWTPs.

Interfacing Machine Learning and Microbial Omics: A Promising Means to Address Environmental Challenges

Microbial communities are ubiquitous and carry an exceptionally broad metabolic capability. Upon environmental perturbation, microbes are also amongst the first natural responsive elements with perturbation-specific cues and markers. These communities are thereby uniquely positioned to inform on the status of environmental conditions. The advent of microbial omics has led to an unprecedented volume of complex microbiological data sets. Importantly, these data sets are rich in biological information with potential for predictive environmental classification and forecasting. However, the patterns in this information are often hidden amongst the inherent complexity of the data. There has been a continued rise in the development and adoption of machine learning (ML) and deep learning architectures for solving research challenges of this sort. Indeed, the interface between molecular microbial ecology and artificial intelligence (AI) appears to show considerable potential for significantly advancing environmental monitoring and management practices through their application. Here, we provide a primer for ML, highlight the notion of retaining biological sample information for supervised ML, discuss workflow considerations, and review the state of the art of the exciting, yet nascent, interdisciplinary field of ML-driven microbial ecology. Current limitations in this sphere of research are also addressed to frame a forward-looking perspective toward the realization of what we anticipate will become a pivotal toolkit for addressing environmental monitoring and management challenges in the years ahead.

Clinical Metagenomics Is Increasingly Accurate and Affordable to Detect Enteric Bacterial Pathogens in Stool

Stool culture is the gold standard method to diagnose enteric bacterial infections; however, many clinical laboratories are transitioning to syndromic multiplex PCR panels. PCR is rapid, accurate, and affordable, yet does not yield subtyping information critical for foodborne disease surveillance. A metagenomics-based stool testing approach could simultaneously provide diagnostic and public health information. Here, we evaluated shotgun metagenomics to assess the detection of common enteric bacterial pathogens in stool. We sequenced 304 stool specimens from 285 patients alongside routine diagnostic testing for Salmonella spp., Campylobacter spp., Shigella spp., and shiga-toxin producing Escherichia coli. Five analytical approaches were assessed for pathogen detection: microbiome profiling, Kraken2, MetaPhlAn, SRST2, and KAT-SECT. Among analysis tools and databases compared, KAT-SECT analysis provided the best sensitivity and specificity for all pathogens tested compared to culture (91.2% and 96.2%, respectively). Where metagenomics detected a pathogen in culture-negative specimens, standard PCR was positive 85% of the time. The cost of metagenomics is approaching the current combined cost of PCR, reflex culture, and whole genome sequencing for pathogen detection and subtyping. As cost, speed, and analytics for single-approach metagenomics improve, it may be more routinely applied in clinical and public health laboratories.

Global mapping of cancers: The Cancer Genome Atlas and beyond

Cancer genomes have been explored from the early 2000s through massive exome sequencing efforts, leading to the publication of The Cancer Genome Atlas in 2013. Sequencing techniques have been developed alongside this project and have allowed scientists to bypass the limitation of costs for whole-genome sequencing (WGS) of single specimens by developing more accurate and extensive cancer sequencing projects, such as deep sequencing of whole genomes and transcriptomic analysis. The Pan-Cancer Analysis of Whole Genomes recently published WGS data from more than 2600 human cancers together with almost 1200 related transcriptomes. The application of WGS on a large database allowed, for the first time in history, a global analysis of features such as molecular signatures, large structural variations and noncoding regions of the genome, as well as the evaluation of RNA alterations in the absence of underlying DNA mutations. The vast amount of data generated still needs to be thoroughly deciphered, and the advent of machine-learning approaches will be the next step towards the generation of personalized approaches for cancer medicine. The present manuscript wants to give a broad perspective on some of the biological evidence derived from the largest sequencing attempts on human cancers so far, discussing advantages and limitations of this approach and its power in the era of machine learning.

Serine and one-carbon metabolisms bring new therapeutic venues in prostate cancer

Serine and one-carbon unit metabolisms are essential biochemical pathways implicated in fundamental cellular functions such as proliferation, biosynthesis of important anabolic precursors and in general for the availability of methyl groups. These two distinct but interacting pathways are now becoming crucial in cancer, the de novo cytosolic serine pathway and the mitochondrial one-carbon metabolism. Apart from their role in physiological conditions, such as epithelial proliferation, the serine metabolism alterations are associated to several highly neoplastic proliferative pathologies. Accordingly, prostate cancer shows a deep rearrangement of its metabolism, driven by the dependency from the androgenic stimulus. Several new experimental evidence describes the role of a few of the enzymes involved in the serine metabolism in prostate cancer pathogenesis. The aim of this study is to analyze gene and protein expression data publicly available from large cancer specimens dataset, in order to further dissect the potential role of the abovementioned metabolism in the complex reshaping of the anabolic environment in this kind of neoplasm. The data suggest a potential role as biomarkers as well as in cancer therapy for the genes (and enzymes) belonging to the one-carbon metabolism in the context of prostatic cancer.

The expression of ELOVL4, repressed by MYCN, defines neuroblastoma patients with good outcome

Cancer cells exhibit dysregulation of critical genes including those involved in lipid biosynthesis, with subsequent defects in metabolism. Here, we show that ELOngation of Very Long chain fatty acids protein 4 (ELOVL4), a rate-limiting enzyme in the biosynthesis of very-long polyunsaturated fatty acids (n-3, ≥28 C), is expressed and transcriptionally repressed by the oncogene MYCN in neuroblastoma cells. In keeping, ELOVL4 positively regulates neuronal differentiation and lipids droplets accumulation in neuroblastoma cells. At the molecular level we found that MYCN binds to the promoter of ELOVL4 in close proximity to the histone deacetylases HDAC1, HDAC2, and the transcription factor Sp1 that can cooperate in the repression of ELOVL4 expression. Accordingly, in vitro differentiation results in an increase of fatty acid with 34 carbons with 6 double bonds (FA34:6); and when MYCN is silenced, FA34:6 metabolite is increased compared with the scrambled. In addition, analysis of large neuroblastoma datasets revealed that ELOVL4 expression is highly expressed in localized clinical stages 1 and 2, and low in high-risk stages 3 and 4. More importantly, high expression of ELOVL4 stratifies a subsets of neuroblastoma patients with good prognosis. Indeed, ELOVL4 expression is a marker of better overall clinical survival also in MYCN not amplified patients and in those with neuroblastoma-associated mutations. In summary, our findings indicate that MYCN, by repressing the expression of ELOVL4 and lipid metabolism, contributes to the progression of neuroblastoma.

NUAK2 and RCan2 participate in the p53 mutant pro-tumorigenic network

Most inactivating mutations in TP53 gene generates neomorphic forms of p53 proteins that experimental evidence and clinical observations suggest to exert gain-of-function effects. While massive effort has been deployed in the dissection of wild type p53 transcriptional programme, p53 mutant pro-tumorigenic gene network is still largely elusive. To help dissecting the molecular basis of p53 mutant GOF, we performed an analysis of a fully annotated genomic and transcriptomic human pancreatic adenocarcinoma to select candidate players of p53 mutant network on the basis their differential expression between p53 mutant and p53 wild-type cohorts and their prognostic value. We identified NUAK2 and RCan2 whose p53 mutant GOF-dependent regulation was further validated in pancreatic cancer cellular model. Our data demonstrated that p53R270H can physically bind RCan2 gene locus in regulatory regions corresponding to the chromatin permissive areas where known binding partners of p53 mutant, such as p63 and Srebp, bind. Overall, starting from clinically relevant data and progressing into experimental validation, our work suggests NUAK2 and RCan2 as novel candidate players of the p53 mutant pro-tumorigenic network whose prognostic and therapeutic interest might attract future studies.

Gut Microbiome and Metabolites in Patients with NAFLD and after Bariatric Surgery: A Comprehensive Review

The prevalence of non-alcoholic fatty liver disease (NAFLD) is increasing, as are other manifestations of metabolic syndrome such as obesity and type 2 diabetes. NAFLD is currently the number one cause of chronic liver disease worldwide. The pathophysiology of NAFLD and disease progression is poorly understood. A potential contributing role for gut microbiome and metabolites in NAFLD is proposed. Currently, bariatric surgery is an effective therapy to prevent the progression of NAFLD and other manifestations of metabolic syndrome such as obesity and type 2 diabetes. This review provides an overview of gut microbiome composition and related metabolites in individuals with NAFLD and after bariatric surgery. Causality remains to be proven. Furthermore, the clinical effects of bariatric surgery on NAFLD are illustrated. Whether the gut microbiome and metabolites contribute to the metabolic improvement and improvement of NAFLD seen after bariatric surgery has not yet been proven. Future microbiome and metabolome research is necessary for elucidating the pathophysiology and underlying metabolic pathways and phenotypes and providing better methods for diagnostics, prognostics and surveillance to optimize clinical care.

Involvement of transcribed lncRNA uc.291 and SWI/SNF complex in cutaneous squamous cell carcinoma

While non-melanoma skin cancers (NMSCs) are the most common tumours in humans, only the sub-type cutaneous squamous cell carcinoma (cSCC), might become metastatic with high lethality. We have recently identified a regulatory pathway involving the lncRNA transcript uc.291 in controlling the expression of epidermal differentiation complex genes via the interaction with ACTL6A, a component of the chromatin remodelling complex SWI/SNF. Since transcribed ultra-conserved regions (T-UCRs) are expressed in normal tissues and are deregulated in tumorigenesis, here we hypothesize a potential role for dysregulation of this axis in cSCC, accounting for the de-differentiation process observed in aggressive poorly differentiated cutaneous carcinomas. We therefore analysed their expression patterns in human tumour biopsies at mRNA and protein levels. The results suggest that by altering chromatin accessibility of the epidermal differentiation complex genes, down-regulation of uc.291 and BRG1 expression contribute to the de-differentiation process seen in keratinocyte malignancy. This provides future direction for the identification of clinical biomarkers in cutaneous SCC. Analysis of publicly available data sets indicates that the above may also be a general feature for SCCs of different origins.

Origin Sample Prediction and Spatial Modeling of Antimicrobial Resistance in Metagenomic Sequencing Data

The steady elaboration of the Metagenomic and Metadesign of Subways and Urban Biomes (MetaSUB) international consortium project raises important new questions about the origin, variation, and antimicrobial resistance of the collected samples. CAMDA (Critical Assessment of Massive Data Analysis, http://camda.info/) forum organizes annual challenges where different bioinformatics and statistical approaches are tested on samples collected around the world for bacterial classification and prediction of geographical origin. This work proposes a method which not only predicts the locations of unknown samples, but also estimates the relative risk of antimicrobial resistance through spatial modeling. We introduce a new component in the standard analysis as we apply a Bayesian spatial convolution model which accounts for spatial structure of the data as defined by the longitude and latitude of the samples and assess the relative risk of antimicrobial resistance taxa across regions which is relevant to public health. We can then use the estimated relative risk as a new measure for antimicrobial resistance. We also compare the performance of several machine learning methods, such as Gradient Boosting Machine, Random Forest, and Neural Network to predict the geographical origin of the mystery samples. All three methods show consistent results with some superiority of Random Forest classifier. In our future work we can consider a broader class of spatial models and incorporate covariates related to the environment and climate profiles of the samples to achieve more reliable estimation of the relative risk related to antimicrobial resistance.