New insights from uncultivated genomes of the global human gut microbiome.

The genome sequences of many species of the human gut microbiome remain unknown, largely owing to challenges in cultivating microorganisms under laboratory conditions. Here we address this problem by reconstructing 60,664 draft prokaryotic genomes from 3,810 faecal metagenomes, from geographically and phenotypically diverse humans. These genomes provide reference points for 2,058 newly identified species-level operational taxonomic units (OTUs), which represents a 50% increase over the previously known phylogenetic diversity of sequenced gut bacteria. On average, the newly identified OTUs comprise 33% of richness and 28% of species abundance per individual, and are enriched in humans from rural populations. A meta-analysis of clinical gut-microbiome studies pinpointed numerous disease associations for the newly identified OTUs, which have the potential to improve predictive models. Finally, our analysis revealed that uncultured gut species have undergone genome reduction that has resulted in the loss of certain biosynthetic pathways, which may offer clues for improving cultivation strategies in the future.

Disentangling direct from indirect relationships in association networks.

Networks are vital tools for understanding and modeling interactions in complex systems in science and engineering, and direct and indirect interactions are pervasive in all types of networks. However, quantitatively disentangling direct and indirect relationships in networks remains a formidable task. Here, we present a framework, called iDIRECT (Inference of Direct and Indirect Relationships with Effective Copula-based Transitivity), for quantitatively inferring direct dependencies in association networks. Using copula-based transitivity, iDIRECT eliminates/ameliorates several challenging mathematical problems, including ill-conditioning, self-looping, and interaction strength overflow. With simulation data as benchmark examples, iDIRECT showed high prediction accuracies. Application of iDIRECT to reconstruct gene regulatory networks in Escherichia coli also revealed considerably higher prediction power than the best-performing approaches in the DREAM5 (Dialogue on Reverse Engineering Assessment and Methods project, #5) Network Inference Challenge. In addition, applying iDIRECT to highly diverse grassland soil microbial communities in response to climate warming showed that the iDIRECT-processed networks were significantly different from the original networks, with considerably fewer nodes, links, and connectivity, but higher relative modularity. Further analysis revealed that the iDIRECT-processed network was more complex under warming than the control and more robust to both random and target species removal (P < 0.001). As a general approach, iDIRECT has great advantages for network inference, and it should be widely applicable to infer direct relationships in association networks across diverse disciplines in science and engineering.

Deciphering the genetic architecture and ethnographic distribution of IRD in three ethnic populations by whole genome sequence analysis.

Patients with inherited retinal dystrophies (IRDs) were recruited from two understudied populations: Mexico and Pakistan as well as a third well-studied population of European Americans to define the genetic architecture of IRD by performing whole-genome sequencing (WGS). Whole-genome analysis was performed on 409 individuals from 108 unrelated pedigrees with IRDs. All patients underwent an ophthalmic evaluation to establish the retinal phenotype. Although the 108 pedigrees in this study had previously been examined for mutations in known IRD genes using a wide range of methodologies including targeted gene(s) or mutation(s) screening, linkage analysis and exome sequencing, the gene mutations responsible for IRD in these 108 pedigrees were not determined. WGS was performed on these pedigrees using Illumina X10 at a minimum of 30X depth. The sequence reads were mapped against hg19 followed by variant calling using GATK. The genome variants were annotated using SnpEff, PolyPhen2, and CADD score; the structural variants (SVs) were called using GenomeSTRiP and LUMPY. We identified potential causative sequence alterations in 61 pedigrees (57%), including 39 novel and 54 reported variants in IRD genes. For 57 of these pedigrees the observed genotype was consistent with the initial clinical diagnosis, the remaining 4 had the clinical diagnosis reclassified based on our findings. In seven pedigrees (12%) we observed atypical causal variants, i.e. unexpected genotype(s), including 4 pedigrees with causal variants in more than one IRD gene within all affected family members, one pedigree with intrafamilial genetic heterogeneity (different affected family members carrying causal variants in different IRD genes), one pedigree carrying a dominant causative variant present in pseudo-recessive form due to consanguinity and one pedigree with a de-novo variant in the affected family member. Combined atypical and large structural variants contributed to about 20% of cases. Among the novel mutations, 75% were detected in Mexican and 50% found in European American pedigrees and have not been reported in any other population while only 20% were detected in Pakistani pedigrees and were not previously reported. The remaining novel IRD causative variants were listed in gnomAD but were found to be very rare and population specific. Mutations in known IRD associated genes contributed to pathology in 63% Mexican, 60% Pakistani and 45% European American pedigrees analyzed. Overall, contribution of known IRD gene variants to disease pathology in these three populations was similar to that observed in other populations worldwide. This study revealed a spectrum of mutations contributing to IRD in three populations, identified a large proportion of novel potentially causative variants that are specific to the corresponding population or not reported in gnomAD and shed light on the genetic architecture of IRD in these diverse global populations.

Effects of Teprotumumab on Eyelid Retraction in Thyroid Eye Disease.

PURPOSE: This study evaluates the efficacy of teprotumumab in reducing eyelid retraction in thyroid eye disease (TED) patients. METHODS: This retrospective study included patients with active or chronic moderate-to-severe TED who completed at least 4 cycles of teprotumumab. Patients with upper and/or lower eyelid retraction, defined as margin-to-reflex distance (MRD) 1 and/or MRD2 of more than 5 mm, in one or OU were included. The main outcome measure was a change in MRD1 and MRD2 after treatment. Changes in MRD1 and MRD2 were each analyzed for correlation (r) with changes in exophthalmolmetry. Student t test was performed for each comparison, and p values <0.05 were considered significant. RESULTS: The study included 91 patients, predominantly female (87%), with an average age of 52.02 ± 14.6 years. The mean baseline proptosis measurement was 21.8 ± 2.9 OD and 21.7 ± 3.3 OS. The average MRD1 was 5.5 ± 1.5 OD and 5.4 ± 1.7 OS, and the average MRD2 was 6.1 ± 1.1 OD and 6.2 ± 1.1 OS. The follow-up duration post-treatment was 37.5 ± 31.7 weeks. At first follow-up post-treatment, the mean change in proptosis, MRD1, and MRD2 were -2.6 ± 2.0 OD, -2.5 ± 2.1 OS, -0.8.5 ± 1.4 OD, -0.8 ± 1.0 OS, and -0.7 ± 0.9 OD, -0.8 ± 1.0 OS, respectively. Correlation analysis showed that proptosis reduction was positively correlated with MRD1 and MRD2 reduction at the first post-treatment follow-up (MRD1: r = 0.23, p value < 0.01; MRD2: r = 0.17, p = 0.03]. CONCLUSION: Teprotumumab treatment improves upper and lower eyelid retraction. The improvement in MRD correlated positively with proptosis reduction, indicating the influence of globe position on eyelid position.

The spatiotemporal gradient of intrusion errors in continuous outcome source memory: Source retrieval is affected by both guessing and intrusions.

Previous research has characterized source retrieval as a thresholded process, which fails on a proportion of trials and leads to guessing, as opposed to a continuous process, in which response precision varies across trials but is never zero. The thresholded view of source retrieval is largely based on the observation of heavy tailed distributions of response errors, thought to reflect a large proportion of "memoryless" trials. In this study, we investigate whether these errors might instead reflect systematic intrusions from other list items which can mimic source guessing. Using the circular diffusion model of decision making, which accounts for both response errors and RTs we found that intrusions account for some, but not all, errors in a continuous-report source memory task. We found that intrusion errors were more likely to come from items studied in nearby locations and times, and were well-described by a spatiotemporal gradient model, but not from semantically or perceptually similar cues. Our findings support a thresholded view of source retrieval but suggest that previous work has overestimated the proportion of guesses which have been conflated with intrusions.

Detection and validation of novel mutations in MERTK in a simplex case of retinal degeneration using WGS and hiPSC-RPEs model.

Inherited retinal degenerations (IRDs) are a group of genetically heterogeneous conditions with a broad phenotypic heterogeneity. Here, we report detection and validation of the underlying cause of progressive retinal degeneration in a nuclear family of European descent with a single affected individual. Whole genome sequencing of the proband and her unaffected sibling identified a novel intron 8 donor splice site variant (c.1296 + 1G>A) and a novel 731 base pair deletion encompassing exon 9 (Chr2:g.112751488_112752218 del) resulting in c.1297_1451del; p.K433_G484fsTer3 in the Mer tyrosine kinase protooncogene (MERTK), which is highly expressed in the retinal pigment epithelium (RPE). The proband carried both variants in the heterozygous state, which segregated with disease in the pedigree. These MERTK variants are predicted to result in the defective splicing of exon 8 and loss of exon 9 respectively. To evaluate the impact of these novel variants, peripheral blood mononuclear cells of the proband and her parents were reprogrammed to humaninduced pluripotent stem cell (hiPSC) lines, which were subsequently differentiated to hiPSC-RPE. Analysis of the proband's hiPSC-RPE revealed the absence of both MERTK transcript and its respective protein as well as abnormal phagocytosis when compared with the parental hiPSC-RPE. In summary, whole genome sequencing identified novel compound heterozygous variants in MERTK as the underlying cause of progressive retinal degeneration in a simplex case. Further, analysis using an hiPSC-RPE model established the functional impact of novel MERTK mutations and revealed the potential mechanism underlying pathology in the proband.

Paper-Based Acoustofluidics for Separating Particles and Cells.

Paper is emerging as a versatile platform for automated fluid handling with a broad range of applications in medical diagnostics and analytical chemistry. However, selectively controlling analyte transport in paper to achieve concentration or selection has been a challenge for functional analysis. Here, by combining paper-based microfluidics with acoustics, we present a rapid and powerful method to size dependently control movement of microparticles and cells in paper using surface acoustic waves (SAW). We demonstrate the unique capability of the paper-based SAW approach to trap and concentrate microparticles in paper and release them when required, achieving collection efficiency of over 98%. Given the correlation between collection efficiency, size, and applied power, the paper-based SAW approach is applied to isolate a mixture of microparticles (1.1, 3.2, and 5 µm in diameter) into different regions and also to trap and concentrate human prostate cancer PC3 cells at a predetermined site. This paper-based SAW approach provides opportunities to develop powerful and low-cost selection and analysis tools, capable of processing complex multicomponent samples, with potential applications in medical diagnostics.

High Concentrations of the Antibiotic Spiramycin in Wastewater Lead to High Abundance of Ammonia-Oxidizing Archaea in Nitrifying Populations.

To evaluate the potential effects of antibiotics on ammonia-oxidizing microbes, multiple tools including quantitative PCR (qPCR), 454-pyrosequencing, and a high-throughput functional gene array (GeoChip) were used to reveal the distribution of ammonia-oxidizing archaea (AOA) and archaeal amoA (Arch-amoA) genes in three wastewater treatment systems receiving spiramycin or oxytetracycline production wastewaters. The qPCR results revealed that the copy number ratios of Arch-amoA to ammonia-oxidizing bacteria (AOB) amoA genes were the highest in the spiramycin full-scale (5.30) and pilot-scale systems (1.49 × 10(-1)), followed by the oxytetracycline system (4.90 × 10(-4)), with no Arch-amoA genes detected in the control systems treating sewage or inosine production wastewater. The pyrosequencing result showed that the relative abundance of AOA affiliated with Thaumarchaeota accounted for 78.5-99.6% of total archaea in the two spiramycin systems, which was in accordance with the qPCR results. Mantel test based on GeoChip data showed that Arch-amoA gene signal intensity correlated with the presence of spiramycin (P < 0.05). Antibiotics explained 25.8% of variations in amoA functional gene structures by variance partitioning analysis. This study revealed the selection of AOA in the presence of high concentrations of spiramycin in activated sludge systems.

Molecular Biomarkers in Ocular Graft-versus-Host Disease: A Systematic Review.

Ocular graft-versus-host disease (oGVHD) affects ~50% of post-stem cell transplant patients and is the only form of GVHD diagnosed without a biopsy. As it must be distinguished from other dry eye diseases, there is a need to identify oGVHD biomarkers to improve diagnosis and treatment. We conducted a systematic review of 19 scholarly articles published from 2018 to 2023 including articles focused on adult patients diagnosed with oGVHD following allogeneic hematopoietic stem cell transplant and used biomarkers as the outcome measure. Articles that were not original investigations or were not published in English were excluded. These clinical investigations explored different molecular oGVHD biomarkers and were identified on 3 October 2023 from the Scopus, PubMed, and Embase databases by using search terms including ocular graft-versus-host disease, biomarkers, cytokines, proteomics, genomics, immune response, imaging techniques, and dry-eye-related key terms. The Newcastle-Ottawa scale for case-control studies was used to assess bias. From the 19 articles included, cytokine, proteomic, lipid, and leukocyte profiles were studied in tear film, as well as ocular surface microbiota and fluorescein staining. Our findings suggest that cytokine profiling is the most studied oGVHD biomarker. Additionally, variations correlating these biomarkers with disease state may lead to a more targeted diagnosis and therapeutic approach. Limitations include language bias, publication bias, and sampling bias, as well as a lack of appropriate controls for included studies.

Retinal blood flow association with age and weight in infants at risk for retinopathy of prematurity.

This prospective study evaluated the relationship between laser speckle contrast imaging (LSCI) ocular blood flow velocity (BFV) and five birth parameters: gestational age (GA), postmenstrual age (PMA), and chronological age (CA) at the time of measurement, birth weight (BW), and current weight (CW) in preterm neonates at risk for retinopathy of prematurity (ROP).38 Neonates with BW < 2 kg, GA < 32 weeks, and PMA between 27-47 weeks underwent 91 LSCI sessions. Correlation tests and regression analysis were performed to quantify relationships between birth parameters and ocular BFV. Mean ocular BFV index in this cohort was 8.8 +/- 4.0 IU. BFV positively correlated with PMA (r = 0.3, p = 0.01), CA (r = 0.3, p = 0.005), and CW (r = 0.3, p = 0.02). BFV did not correlate with GA nor BW (r=-0.2 and r=-0.05, p > 0.05). Regression analysis with mixed models demonstrated that BFV increased by 1.2 for every kilogram of CW, by 0.34 for every week of CA, and by 0.36 for every week of PMA (p = 0.03, 0.004, 0.007, respectively). Our findings indicate that increased age and weight are associated with increased ocular BFV measured using LSCI in premature infants. Future studies investigating the associations between ocular BFV and ROP clinical severity must control for age and/or weight of the infant.

Retinal blood flow association with age and weight in infants at risk for retinopathy of prematurity.

This prospective study evaluated the relationship between laser speckle contrast imaging (LSCI) ocular blood flow velocity (BFV) and five birth parameters: gestational age (GA), postmenstrual age (PMA) and chronological age (CA) at the time of measurement, birth weight (BW), and current weight (CW) in preterm neonates at risk for retinopathy of prematurity (ROP). 38 Neonates with BW < 2 kg, GA < 32 weeks, and PMA between 27 and 47 weeks underwent 91 LSCI sessions. Correlation tests and regression analysis were performed to quantify relationships between birth parameters and ocular BFV. Mean ocular BFV index in this cohort was 8.8 +/- 4.0 IU. BFV positively correlated with PMA (r = 0.3, p = 0.01), CA (r = 0.3, p = 0.005), and CW (r = 0.3, p = 0.02). BFV did not correlate with GA nor BW (r = - 0.2 and r = - 0.05, p > 0.05). Regression analysis with mixed models demonstrated that BFV increased by 1.2 for every kilogram of CW, by 0.34 for every week of CA, and by 0.36 for every week of PMA (p = 0.03, 0.004, 0.007, respectively). Our findings indicate that increased age and weight are associated with increased ocular BFV measured using LSCI in premature infants. Future studies investigating the associations between ocular BFV and ROP clinical severity must control for age and/or weight of the infant.

Warming effects on grassland soil microbial communities are amplified in cool months.

Global warming modulates soil respiration (RS) via microbial decomposition, which is seasonally dependent. Yet, the magnitude and direction of this modulation remain unclear, partly owing to the lack of knowledge on how microorganisms respond to seasonal changes. Here, we investigated the temporal dynamics of soil microbial communities over 12 consecutive months under experimental warming in a tallgrass prairie ecosystem. The interplay between warming and time altered (P < 0.05) the taxonomic and functional compositions of microbial communities. During the cool months (January to February and October to December), warming induced a soil microbiome with a higher genomic potential for carbon decomposition, community-level ribosomal RNA operon (rrn) copy numbers, and microbial metabolic quotients, suggesting that warming stimulated fast-growing microorganisms that enhanced carbon decomposition. Modeling analyses further showed that warming reduced the temperature sensitivity of microbial carbon use efficiency (CUE) by 28.7% when monthly average temperature was low, resulting in lower microbial CUE and higher heterotrophic respiration (Rh) potentials. Structural equation modeling showed that warming modulated both Rh and RS directly by altering soil temperature and indirectly by influencing microbial community traits, soil moisture, nitrate content, soil pH, and gross primary productivity. The modulation of Rh by warming was more pronounced in cooler months compared to warmer ones. Together, our findings reveal distinct warming-induced effects on microbial functional traits in cool months, challenging the norm of soil sampling only in the peak growing season, and advancing our mechanistic understanding of the seasonal pattern of RS and Rh sensitivity to warming.

Identifying species-specific k-mers for fast and accurate metagenotyping with Maast and GT-Pro.

Genotyping single-nucleotide polymorphisms (SNPs) in microbiomes enables strain-level quantification. In this protocol, we describe a computational pipeline that performs fast and accurate SNP genotyping using metagenomic data. We first demonstrate how to use Maast to catalog SNPs from microbial genomes. Then we use GT-Pro to extract unique SNP-covering k-mers, optimize a data structure for storing these k-mers, and finally perform metagenotyping. For proof of concept, the protocol leverages public whole-genome sequences to metagenotype a synthetic community. For complete details on the use and execution of this protocol, please refer to Shi et al. (2022a)1 and Shi et al. (2022b).2.

Maast: genotyping thousands of microbial strains efficiently.

Existing single nucleotide polymorphism (SNP) genotyping algorithms do not scale for species with thousands of sequenced strains, nor do they account for conspecific redundancy. Here we present a bioinformatics tool, Maast, which empowers population genetic meta-analysis of microbes at an unrivaled scale. Maast implements a novel algorithm to heuristically identify a minimal set of diverse conspecific genomes, then constructs a reliable SNP panel for each species, and enables rapid and accurate genotyping using a hybrid of whole-genome alignment and k-mer exact matching. We demonstrate Maast's utility by genotyping thousands of Helicobacter pylori strains and tracking SARS-CoV-2 diversification.

Pitfalls of genotyping microbial communities with rapidly growing genome collections.

Detecting genetic variants in metagenomic data is a priority for understanding the evolution, ecology, and functional characteristics of microbial communities. Many tools that perform this metagenotyping rely on aligning reads of unknown origin to a database of sequences from many species before calling variants. In this synthesis, we investigate how databases of increasingly diverse and closely related species have pushed the limits of current alignment algorithms, thereby degrading the performance of metagenotyping tools. We identify multi-mapping reads as a prevalent source of errors and illustrate a trade-off between retaining correct alignments versus limiting incorrect alignments, many of which map reads to the wrong species. Then we evaluate several actionable mitigation strategies and review emerging methods showing promise to further improve metagenotyping in response to the rapid growth in genome collections. Our results have implications beyond metagenotyping to the many tools in microbial genomics that depend upon accurate read mapping.

Fast and accurate metagenotyping of the human gut microbiome with GT-Pro.

Single nucleotide polymorphisms (SNPs) in metagenomics are used to quantify population structure, track strains and identify genetic determinants of microbial phenotypes. However, existing alignment-based approaches for metagenomic SNP detection require high-performance computing and enough read coverage to distinguish SNPs from sequencing errors. To address these issues, we developed the GenoTyper for Prokaryotes (GT-Pro), a suite of methods to catalog SNPs from genomes and use unique k-mers to rapidly genotype these SNPs from metagenomes. Compared to methods that use read alignment, GT-Pro is more accurate and two orders of magnitude faster. Using high-quality genomes, we constructed a catalog of 104 million SNPs in 909 human gut species and used unique k-mers targeting this catalog to characterize the global population structure of gut microbes from 7,459 samples. GT-Pro enables fast and memory-efficient metagenotyping of millions of SNPs on a personal computer.

Scalable Microbial Strain Inference in Metagenomic Data Using StrainFacts.

While genome databases are nearing a complete catalog of species commonly inhabiting the human gut, their representation of intraspecific diversity is lacking for all but the most abundant and frequently studied taxa. Statistical deconvolution of allele frequencies from shotgun metagenomic data into strain genotypes and relative abundances is a promising approach, but existing methods are limited by computational scalability. Here we introduce StrainFacts, a method for strain deconvolution that enables inference across tens of thousands of metagenomes. We harness a "fuzzy" genotype approximation that makes the underlying graphical model fully differentiable, unlike existing methods. This allows parameter estimates to be optimized with gradient-based methods, speeding up model fitting by two orders of magnitude. A GPU implementation provides additional scalability. Extensive simulations show that StrainFacts can perform strain inference on thousands of metagenomes and has comparable accuracy to more computationally intensive tools. We further validate our strain inferences using single-cell genomic sequencing from a human stool sample. Applying StrainFacts to a collection of more than 10,000 publicly available human stool metagenomes, we quantify patterns of strain diversity, biogeography, and linkage-disequilibrium that agree with and expand on what is known based on existing reference genomes. StrainFacts paves the way for large-scale biogeography and population genetic studies of microbiomes using metagenomic data.

The green tea polyphenol (-)-epigallocatechin-3-gallate ameliorates experimental immune-mediated glomerulonephritis.

The unchecked overproduction of reactive oxygen and nitrogen species by inflammatory cells can cause tissue damage, intensify inflammation, promote apoptosis, and accelerate the progression of immune-mediated glomerulonephritis (GN). Here we tested whether the anti-inflammatory and antioxidant properties of the green tea polyphenol (-)-epigallocatechin-3-gallate (EGCG) favorably affect the development of immune-mediated GN. Pretreatment of 129/svJ mice with EGCG from 2 days before to 2 weeks after the induction of GN led to reduced proteinuria and serum creatinine, and marked improvement in renal histology when compared with vehicle-pretreated diseased mice. This pretreatment reduced oxidative stress, and normalized osteopontin, p65/nuclear factor-κB, inducible nitric oxide synthase, nitric oxide metabolites, p-Akt, phosphorylated extracellular signal-regulated kinases 1 and 2, p47phox, and myeloperoxidase, all of which were elevated in vehicle-pretreated diseased mice. Levels of glutathione peroxidase and peroxisome proliferator-activated receptor-γ (PPARγ), both reduced in the vehicle-pretreated diseased mice, were normalized. This renoprotective effect was reversed by concomitant administration of the PPARγ antagonist GW9662 throughout the EGCG pretreatment period. Importantly, mortality and renal dysfunction were significantly attenuated even when the polyphenol treatment was initiated 1 week after the onset of GN. Thus, EGCG reversed the progression of immune-mediated GN in mice by targeting redox and inflammatory pathways.

A circular diffusion model of continuous-outcome source memory retrieval: Contrasting continuous and threshold accounts.

A circular analogue of the diffusion model adapted for continuous response tasks is applied to a continuous-outcome source memory task. In contrast to existing models of source retrieval that attribute all of the variability in responding to memory, the circular diffusion model decomposes noise into variability arising from memory and from decision processes. We compared three models: (1) a single diffusion process with trial-to-trial variability in drift rate, (2) a mixture of two diffusion processes, one with positive drift that does not vary from trial-to-trial, and a second zero-drift process that represents discrete guessing, and (3) a hybrid model that mixed positive and zero-drift processes with trial-to-trial variability in the positive drift process. Comparison of model fits to joint response error and response-time (RT) data suggest that a memory strength threshold under which no information is retrieved appears to underlie responding in a continuous-report source memory task. Additionally, we also conditioned participants' source responding on their confidence in an old/new recognition task, ruling out the possibility that participant guessing was only due to unrecognized items. Overall, our findings support an all-or-none or some-or none view of source memory retrieval and pose a challenge to continuous models of source memory.

A unified catalog of 204,938 reference genomes from the human gut microbiome.

Comprehensive, high-quality reference genomes are required for functional characterization and taxonomic assignment of the human gut microbiota. We present the Unified Human Gastrointestinal Genome (UHGG) collection, comprising 204,938 nonredundant genomes from 4,644 gut prokaryotes. These genomes encode >170 million protein sequences, which we collated in the Unified Human Gastrointestinal Protein (UHGP) catalog. The UHGP more than doubles the number of gut proteins in comparison to those present in the Integrated Gene Catalog. More than 70% of the UHGG species lack cultured representatives, and 40% of the UHGP lack functional annotations. Intraspecies genomic variation analyses revealed a large reservoir of accessory genes and single-nucleotide variants, many of which are specific to individual human populations. The UHGG and UHGP collections will enable studies linking genotypes to phenotypes in the human gut microbiome.