Zero-Inflated gaussian mixed models for analyzing longitudinal microbiome data

Zero inflated high dimensional compositional data with DeepInsight

Through the Human Microbiome Project, research on human-associated microbiomes has been conducted in various fields. New sequencing techniques such as Next Generation Sequencing (NGS) and High-Throughput Sequencing (HTS) have enabled the inclusion of a wide range of features of the microbiome. These advancements have also contributed to the development of numerical proxies like Operational Taxonomic Units (OTUs) and Amplicon Sequence Variants (ASVs). Studies involving such microbiome data often encounter zero-inflated and high-dimensional problems. Based on the need to address these two issues and the recent emphasis on compositional interpretation of microbiome data, we conducted our research. To solve the zero-inflated problem in compositional microbiome data, we transformed the data onto the surface of the hypersphere using a square root transformation. Then, to solve the high-dimensional problem, we modified DeepInsight, an image-generating method using Convolutional Neural Networks (CNNs), to fit the hypersphere space. Furthermore, to resolve the common issue of distinguishing between true zero values and fake zero values in zero-inflated images, we added a small value to the true zero values. We validated our approach using pediatric inflammatory bowel disease (IBD) fecal sample data and achieved an area under the curve (AUC) value of 0.847, which is higher than the previous study's result of 0.83.

The effect of pertussis vaccination in pregnancy on the immunogenicity of acellular or whole-cell pertussis vaccination in Gambian infants (GaPS): a single-centre, randomised, controlled, double-blind, phase 4 trial

BACKGROUND

Vaccinating women against pertussis in pregnancy protects young infants from severe disease and death. Vaccination-induced maternally derived antibodies, however, might subsequently modulate (and specifically blunt) the infant's serological response to their primary series of pertussis vaccinations. We examined the effect of pertussis immunisation in pregnancy on the immunogenicity of primary acellular or whole-cell pertussis vaccines in a west African cohort.

METHODS

GaPs was a randomised, controlled, double-blind, phase 4 trial conducted in The Gambia. We used a predefined block randomisation scheme to randomly assign healthy, HIV-negative, pregnant participants (1:1) to receive a pertussis-containing (tetanus-diphtheria-acellular pertussis-inactivated polio virus [Tdap-IPV]) or tetanus-toxoid only vaccine at 28-34 weeks' gestation. At the same time, their infants were randomly assigned (1:1) to receive diphtheria-tetanus-acellular pertussis (DTaP) or diphtheria-tetanus-whole-cell pertussis (DTwP) primary vaccine at 8, 12, and 16 weeks postnatally. Participants and trial staff were masked to the allocation of the maternal vaccine. The field team and participants became unmasked to the allocation of the infant vaccine at 16 weeks; laboratory staff and all other investigators remained masked to infant vaccine allocation until the end of the trial. The primary outcome was geometric mean concentration (GMC) of infant pertussis toxin-specific antibodies at 20 weeks and 9 months postnatally and was assessed in infants who received all three doses of the primary vaccine. Secondary outcomes included memory B-cell responses, and exploratory outcomes were total pertussis-specific antibody binding concentrations and functional antibody titres (pertussis toxin-specific neutralising activity [PTNA] and serum bactericidal activity [SBA]). Vaccine reactogenicity was assessed in mothers and infants for 3 days after each vaccine dose. Pregnant women had an extra safety visit 7 days after vaccination. The study is registered with ClinicalTrials.gov, NCT03606096.

FINDINGS

Between Feb 13, 2019, and May 17, 2021, we enrolled 343 maternal-infant pairs. 239 (77%) infants were included in the per-protocol immunogenicity analysis. Among infants of mothers receiving Tdap-IPV in pregnancy, at 20 weeks postnatally, the GMCs of anti-pertussis toxin IgG were more than three-fold lower in infants vaccinated with three doses of DTwP (n=64) than in infants vaccinated with three doses of DTaP (n=53; adjusted geometric mean ratio 0·28, 98·75% CI 0·16-0·50). This difference persisted up to 9 months (0·31, 0·17-0·55). Conversely, among infants born to tetanus toxoid-immunised mothers, post-vaccination GMCs of anti-pertussis toxin IgG at 9 months were higher in those vaccinated with DTwP (n=58) than in those vaccinated with DTaP (n=64; 2·02, 1·15-3·55). Tdap-IPV immunisation in pregnancy blunted anti-pertussis toxin IgG following primary vaccination in all infants but particularly in those receiving DTwP, with GMCs of anti-pertussis toxin IgG more than eight-fold lower in DTwP-vaccinated infants born to Tdap-IPV-vaccinated mothers than in DTwP-vaccinated infants born to tetanus toxoid-immunised mothers (0·12, 98·75% CI 0·07-0·22 at 20 weeks; 0·07, 0·03-0·17 at 9 months). Similarly, DTwP-vaccinated infants born to Tdap-IPV-vaccinated mothers also showed significant blunting of PTNA, SBA, total pertussis-specific antibody binding, and memory B-cell responses after primary immunisation, whereas minimal blunting was observed among DTaP-vaccinated infants. However, the absolute levels of these responses generated by DTwP-vaccinated infants remained similar to or, in many cases, were higher than those generated by DTaP-vaccinated infants. There was no difference in reactogenicity between the two maternal vaccines, with most reactions graded 0 or 1. There were no serious adverse events related to vaccination or trial participation.

INTERPRETATION

Vaccinating women with Tdap-IPV in pregnancy was safe and well tolerated in a sub-Saharan African setting and boosted the quantity and quality of pertussis-specific antibodies in infants in early life. Although Tdap-IPV was associated with relative blunting of the immune response to the DTwP primary vaccination series, pertussis-specific antibody quality and memory B-cell responses were nevertheless preserved, regardless of the vaccine given during pregnancy.

FUNDING

GaPs was conducted as part of the Pertussis Correlates Of Protection Europe (PERISCOPE) consortium, which received funding from the Innovative Medicines Initiative 2 Joint Undertaking under grant agreement 115910. This Joint Undertaking receives support from the EU's Horizon 2020 research and innovation programme, the European Federation of Pharmaceutical Industries and Associations, and the Bill & Melinda Gates Foundation.

Microbial network inference for longitudinal microbiome studies with LUPINE

BACKGROUND

The microbiome is a complex ecosystem of interdependent taxa that has traditionally been studied through cross-sectional studies. However, longitudinal microbiome studies are becoming increasingly popular. These studies enable researchers to infer taxa associations towards the understanding of coexistence, competition, and collaboration between microbes across time. Traditional metrics for association analysis, such as correlation, are limited due to the data characteristics of microbiome data (sparse, compositional, multivariate). Several network inference methods have been proposed, but have been largely unexplored in a longitudinal setting.

RESULTS

We introduce LUPINE (LongitUdinal modelling with Partial least squares regression for NEtwork inference), a novel approach that leverages on conditional independence and low-dimensional data representation. This method is specifically designed to handle scenarios with small sample sizes and small number of time points. LUPINE is the first method of its kind to infer microbial networks across time, while considering information from all past time points and is thus able to capture dynamic microbial interactions that evolve over time. We validate LUPINE and its variant, LUPINE_single (for single time point analysis) in simulated data and four case studies, where we highlight LUPINE's ability to identify relevant taxa in each study context, across different experimental designs (mouse and human studies, with or without interventions, and short or long time courses). To detect changes in the networks across time and groups or in response to external disturbances, we used different metrics to compare the inferred networks.

CONCLUSIONS

LUPINE is a simple yet innovative network inference methodology that is suitable for, but not limited to, analysing longitudinal microbiome data. The R code and data are publicly available for readers interested in applying these new methods to their studies. Video Abstract.

Examining emotion reactivity to politically polarizing media in a randomized controlled trial of mindfulness training versus active coping training

Emotional appraisals of political stimuli (e.g., videos) have been shown to drive shared neural encoding, which correspond to shared, yet divisive, interpretations of such stimuli. However, mindfulness practice may entrain a form of emotion regulation that de-automatizes social biases, possibly through alteration of such neural mechanisms. The present study combined a naturalistic neuroimaging paradigm and a randomized controlled trial to examine the effects of short-term mindfulness training (MT) (n = 35) vs structurally equivalent Cognitive Reappraisal training (CT) (n = 37) on politically-situated emotions while evaluating the mechanistic role of prefrontal cortical neural synchrony. Participants underwent functional near-infrared spectroscopy (fNIRS) recording while viewing inflammatory partisan news clips and continuously rating their momentary discrete emotions. MT participants were more likely to respond with extreme levels of anger (odds ratio = 0.12, p < 0.001) and disgust (odds ratio = 0.08, p < 0.001) relative to CT participants. Neural synchrony-based analyses suggested that participants with extreme emotion reactions exhibited greater prefrontal cortical neural synchrony, but that this pattern was less prominent in participants receiving MT relative to CT (CT > MT; channel 1 ISC = 0.040, p = 0.030).

The impact of regular sauerkraut consumption on the human gut microbiota: a crossover intervention trial

BACKGROUND

Sauerkraut is a fermented food that has been suspected to have a beneficial impact on the gut microbiome, but scientific evidence is sparse. In this crossover intervention trial with 87 participants (DRKS00027007), we investigated the impact of daily consumption of fresh or pasteurized sauerkraut for 4 weeks on gut microbial composition and the metabolome in a healthy study population.

RESULTS

Using shotgun metagenomic sequencing, we observed changes in single bacterial species following fresh and pasteurized sauerkraut consumption. More pronounced changes were observed in the pasteurized sauerkraut intervention arm. Only pasteurized sauerkraut consumption increased serum short-chain fatty acids (SCFAs).

CONCLUSIONS

The gut microbiome of healthy individuals is rather resilient to short-term dietary interventions even though single species might be affected by sauerkraut consumption. Video Abstract.

Association of body index with fecal microbiome in children cohorts with ethnic-geographic factor interaction: accurately using a Bayesian zero-inflated negative binomial regression model

The exponential growth of high-throughput sequencing (HTS) data on the microbial communities presents researchers with an unparalleled opportunity to delve deeper into the association of microorganisms with host phenotype. However, this growth also poses a challenge, as microbial data are complex, sparse, discrete, and prone to zero inflation. Herein, by utilizing 10 distinct counting models for analyzing simulated data, we proposed an innovative Bayesian zero-inflated negative binomial (ZINB) regression model that is capable of identifying differentially abundant taxa associated with distinctive host phenotypes and quantifying the effects of covariates on these taxa. Our proposed model exhibits excellent accuracy compared with conventional Hurdle and INLA models, especially in scenarios characterized by inflation and overdispersion. Moreover, we confirm that dispersion parameters significantly affect the accuracy of model results, with defects gradually alleviating as the number of analyzed samples increases. Subsequently applying our model to amplicon data in real multi-ethnic children cohort, we found that only a subset of taxa were identified as having zero inflation in real data, suggesting that the prevailing understanding and processing of microbial count data in most previous microbiome studies were overly dogmatic. In practice, our pipeline of integrating bacterial differential abundance in microbiome data and relevant covariates is effective and feasible. Taken together, our method is expected to be extended to the microbiota studies of various multi-cohort populations.

IMPORTANCE

The microbiome is closely associated with physical indicators of the body, such as height, weight, age and BMI, which can be used as measures of human health. Accurately identifying which taxa in the microbiome are closely related to indicators of physical development is valuable as microbial markers of regional child growth trajectory. Zero-inflated negative binomial (ZINB) model, a type of Bayesian generalized linear model, can be effectively modeled in complex biological systems. We present an innovative ZINB regression model that is capable of identifying differentially abundant taxa associated with distinctive host phenotypes and quantifying the effects of covariates on these taxa, and demonstrate that its accuracy is superior to traditional Hurdle and INLA models. Our pipeline of integrating bacterial differential abundance in microbiome data and relevant covariates is effective and feasible.

Genetic remodeling of soil diazotrophs enables partial replacement of synthetic nitrogen fertilizer with biological nitrogen fixation in maize

Increasing biological nitrogen (N) fixation (BNF) in maize production could reduce the environmental impacts of N fertilizer use, but reactive N in the rhizosphere of maize limits the BNF process. Using non-transgenic methods, we developed gene-edited strains of Klebsiella variicola (Kv137-2253) and Kosakonia sacchari (Ks6-5687) bacteria optimized for root-associated BNF and ammonium excretion in N-rich conditions. The aim of this research was to elucidate the mechanism of action of these strains. We present evidence from in vitro, in planta and field experiments that confirms that our genetic remodeling strategy derepresses BNF activity in N-rich systems and increases ammonium excretion by orders of magnitude above the respective wildtype strains. BNF is demonstrated in controlled environments by the transfer of labeled 15N2 gas from the rhizosphere to the chlorophyll of inoculated maize plants. This was corroborated in several 15N isotope tracer field experiments where inoculation with the formulated, commercial-grade product derived from the gene-edited strains (PIVOT BIO PROVEN® 40) provided on average 21 kg N ha-1 to the plant by the VT-R1 growth stages. Data from small-plot and on-farm trials suggest that this technology can improve crop N status pre-flowering and has potential to mitigate the risk of yield loss associated with a reduction in synthetic N fertilizer inputs.

A comparative analysis of mutual information methods for pairwise relationship detection in metagenomic data

BACKGROUND

Construction of co-occurrence networks in metagenomic data often employs correlation to infer pairwise relationships between microbes. However, biological systems are complex and often display qualities non-linear in nature. Therefore, the reliance on correlation alone may overlook important relationships and fail to capture the full breadth of intricacies presented in underlying interaction networks. It is of interest to incorporate metrics that are not only robust in detecting linear relationships, but non-linear ones as well.

RESULTS

In this paper, we explore the use of various mutual information (MI) estimation approaches for quantifying pairwise relationships in biological data and compare their performances against two traditional measures-Pearson's correlation coefficient, r, and Spearman's rank correlation coefficient, ρ. Metrics are tested on both simulated data designed to mimic pairwise relationships that may be found in ecological systems and real data from a previous study on C. diff infection. The results demonstrate that, in the case of asymmetric relationships, mutual information estimators can provide better detection ability than Pearson's or Spearman's correlation coefficients. Specifically, we find that these estimators have elevated performances in the detection of exploitative relationships, demonstrating the potential benefit of including them in future metagenomic studies.

CONCLUSIONS

Mutual information (MI) can uncover complex pairwise relationships in biological data that may be missed by traditional measures of association. The inclusion of such relationships when constructing co-occurrence networks can result in a more comprehensive analysis than the use of correlation alone.

Current Trends and Challenges of Microbiome Research in Prostate Cancer

PURPOSE OF REVIEW

The role of the gut microbiome in prostate cancer is an emerging area of research interest. However, no single causative organism has yet been identified. The goal of this paper is to examine the role of the microbiome in prostate cancer and summarize the challenges relating to methodology in specimen collection, sequencing technology, and interpretation of results.

RECENT FINDINGS

Significant heterogeneity still exists in methodology for stool sampling/storage, preservative options, DNA extraction, and sequencing database selection/in silico processing. Debate persists over primer choice in amplicon sequencing as well as optimal methods for data normalization. Statistical methods for longitudinal microbiome analysis continue to undergo refinement. While standardization of methodology may help yield more consistent results for organism identification in prostate cancer, this is a difficult task due to considerable procedural variation at each step in the process. Further reproducibility and methodology research is required.

Managing emotions in the age of political polarization: A randomized controlled trial comparing mindfulness to cognitive reappraisal

Emotional appraisals of political stimuli (e.g., videos) have been shown to drive shared neural encoding, which correspond to shared, yet divisive, interpretations of such stimuli. However, mindfulness practice may entrain a form of emotion regulation that de-automatizes social biases, possibly through alteration of such neural mechanisms. The present study combined a naturalistic neuroimaging paradigm and a randomized controlled trial to examine the effects of short-term mindfulness training (MT) (n = 35) vs structurally equivalent Cognitive Reappraisal training (CT) (n = 37) on politically-situated emotions while evaluating the mechanistic role of prefrontal cortical neural synchrony. Participants underwent functional near-infrared spectroscopy (fNIRS) recording while viewing inflammatory partisan news clips and continuously rating their momentary discrete emotions. MT participants were more likely to respond with extreme levels of anger (odds ratio = 0.12, p < .001) and disgust (odds ratio = 0.08, p < .001) relative to CT participants. Neural synchrony-based analyses suggested that participants with extreme emotion reactions exhibited greater prefrontal cortical neural synchrony, but that this pattern was less prominent in participants receiving MT relative to CT (CT > MT; channel 1 ISC = .040, p = .030).

Discovery of Bacterial Key Genes from 16S rRNA-Seq Profiles That Are Associated with the Complications of SARS-CoV-2 Infections and Provide Therapeutic Indications

SARS-CoV-2 infections, commonly referred to as COVID-19, remain a critical risk to both human life and global economies. Particularly, COVID-19 patients with weak immunity may suffer from different complications due to the bacterial co-infections/super-infections/secondary infections. Therefore, different variants of alternative antibacterial therapeutic agents are required to inhibit those infection-causing drug-resistant pathogenic bacteria. This study attempted to explore these bacterial pathogens and their inhibitors by using integrated statistical and bioinformatics approaches. By analyzing bacterial 16S rRNA sequence profiles, at first, we detected five bacterial genera and taxa (Bacteroides, Parabacteroides, Prevotella Clostridium, Atopobium, and Peptostreptococcus) based on differentially abundant bacteria between SARS-CoV-2 infection and control samples that are significantly enriched in 23 metabolic pathways. A total of 183 bacterial genes were found in the enriched pathways. Then, the top-ranked 10 bacterial genes (accB, ftsB, glyQ, hldD, lpxC, lptD, mlaA, ppsA, ppc, and tamB) were selected as the pathogenic bacterial key genes (bKGs) by their protein-protein interaction (PPI) network analysis. Then, we detected bKG-guided top-ranked eight drug molecules (Bemcentinib, Ledipasvir, Velpatasvir, Tirilazad, Acetyldigitoxin, Entreatinib, Digitoxin, and Elbasvir) by molecular docking. Finally, the binding stability of the top-ranked three drug molecules (Bemcentinib, Ledipasvir, and Velpatasvir) against three receptors (hldD, mlaA, and lptD) was investigated by computing their binding free energies with molecular dynamic (MD) simulation-based MM-PBSA techniques, respectively, and was found to be stable. Therefore, the findings of this study could be useful resources for developing a proper treatment plan against bacterial co-/super-/secondary-infection in SARS-CoV-2 infections.

Infarct-related structural disconnection and delirium in surgical aortic valve replacement patients

OBJECTIVE

Although acute brain infarcts are common after surgical aortic valve replacement (SAVR), they are often unassociated with clinical stroke symptoms. The relationship between clinically "silent" infarcts and in-hospital delirium remains uncertain; obscured, in part, by how infarcts have been traditionally summarized as global metrics, independent of location or structural consequence. We sought to determine if infarct location and related structural connectivity changes were associated with postoperative delirium after SAVR.

METHODS

A secondary analysis of a randomized multicenter SAVR trial of embolic protection devices (NCT02389894) was conducted, excluding participants with clinical stroke or incomplete neuroimaging (N = 298; 39% female, 7% non-White, 74 ± 7 years). Delirium during in-hospital recovery was serially screened using the Confusion Assessment Method. Parcellation and tractography atlas-based neuroimaging methods were used to determine infarct locations and cortical connectivity effects. Mixed-effect, zero-inflated gaussian modeling analyses, accounting for brain region-specific infarct characteristics, were conducted to examine for differences within and between groups by delirium status and perioperative neuroprotection device strategy.

RESULTS

23.5% participants experienced postoperative delirium. Delirium was associated with significantly increased lesion volumes in the right cerebellum and temporal lobe white matter, while diffusion weighted imaging infarct-related structural disconnection (DWI-ISD) was observed in frontal and temporal lobe regions (p-FDR < 0.05). Fewer brain regions demonstrated DWI-ISD loss in the suction-based neuroprotection device group, relative to filtration-based device or standard aortic cannula.

INTERPRETATION

Structural disconnection from acute infarcts was greater in patients who experienced postoperative delirium, suggesting that the impact from covert perioperative infarcts may not be as clinically "silent" as commonly assumed.

Bile acid metabolites predict multiple sclerosis progression and supplementation is safe in progressive disease

Background

Bile acid metabolism is altered in multiple sclerosis (MS) and tauroursodeoxycholic acid (TUDCA) supplementation ameliorated disease in mouse models of MS.

Methods

Global metabolomics was performed in an observational cohort of people with MS followed by pathway analysis to examine relationships between baseline metabolite levels and subsequent brain and retinal atrophy. A double-blind, placebo-controlled trial, was completed in people with progressive MS (PMS), randomized to receive either TUDCA (2g daily) or placebo for 16 weeks. Participants were followed with serial clinical and laboratory assessments. Primary outcomes were safety and tolerability of TUDCA, and exploratory outcomes included changes in clinical, laboratory and gut microbiome parameters.

Results

In the observational cohort, higher primary bile acid levels at baseline predicted slower whole brain, brain substructure and specific retinal layer atrophy. In the clinical trial, 47 participants were included in our analyses (21 in placebo arm, 26 in TUDCA arm). Adverse events did not significantly differ between arms (p=0.77). The TUDCA arm demonstrated increased serum levels of multiple bile acids. No significant differences were noted in clinical or fluid biomarker outcomes. Central memory CD4+ and Th1/17 cells decreased, while CD4+ naïve cells increased in the TUDCA arm compared to placebo. Changes in the composition and function of gut microbiota were also noted in the TUDCA arm compared to placebo.

Conclusion

Bile acid metabolism in MS is linked with brain and retinal atrophy. TUDCA supplementation in PMS is safe, tolerable and has measurable biological effects that warrant further evaluation in larger trials with a longer treatment duration.

Functional Trait Responses of Sophora alopecuroides L. Seedlings to Diverse Environmental Stresses in the Desert Steppe of Ningxia, China

The seedling stage of plants is a crucial and vulnerable period in population and community dynamics. Despite this, studies on how plant traits respond to different environmental stresses often tend to overlook this early stage. Our study focused on Sophora alopecuroides L. seedlings in Ningxia Yanchi desert steppe, analyzing the effects of sand burial, salinity, and drought on their key aboveground and belowground traits. The results showed that sand burial significantly negatively affected stem biomass (SB), leaf biomass (LB), stem diameter (SD), leaf length (LL), leaf width (LW), leaf area (LA), and total root volume (RV), but positively influenced total root length (RL). As sand burial depth increased, SB, LB, SD, LL, LW, LA, RV, root biomass (RB), RV, and lateral root numbers (LRN) significantly decreased. Salinity stress negatively affected SB, LB, SD, LL, LW, LA, RB, RL, and RV, with these traits declining as the stress concentration increased. Drought stress had a positive effect on SD and LL, with both traits showing an increase as the intensity of the drought stress intensified; however, it adversely affected RL. In Ningxia Yanchi desert steppe, salinity stress had the most significant effect on the traits of S. alopecuroides seedlings, followed by sand burial, with drought having the least significant effect. This study provides essential theoretical support for understanding how S. alopecuroides seedlings cope with environmental stresses in their early life stages.

The upper respiratory tract microbiota of healthy adults is affected by Streptococcus pneumoniae carriage, smoking habits, and contact with children

BACKGROUND

The microbiota of the upper respiratory tract is increasingly recognized as a gatekeeper of respiratory health. Despite this, the microbiota of healthy adults remains understudied. To address this gap, we investigated the composition of the nasopharyngeal and oropharyngeal microbiota of healthy adults, focusing on the effect of Streptococcus pneumoniae carriage, smoking habits, and contact with children.

RESULTS

Differential abundance analysis indicated that the microbiota of the oropharynx was significantly different from that of the nasopharynx (P < 0.001) and highly discriminated by a balance between the classes Negativicutes and Bacilli (AUC of 0.979). Moreover, the oropharynx was associated with a more homogeneous microbiota across individuals, with just two vs. five clusters identified in the nasopharynx. We observed a shift in the nasopharyngeal microbiota of carriers vs. noncarriers with an increased relative abundance of Streptococcus, which summed up to 30% vs. 10% in noncarriers and was not mirrored in the oropharynx. The oropharyngeal microbiota of smokers had a lower diversity than the microbiota of nonsmokers, while no differences were observed in the nasopharyngeal microbiota. In particular, the microbiota of smokers, compared with nonsmokers, was enriched (on average 16-fold) in potential pathogenic taxa involved in periodontal diseases of the genera Bacillus and Burkholderia previously identified in metagenomic studies of cigarettes. The microbiota of adults with contact with children resembled the microbiota of children. Specifically, the nasopharyngeal microbiota of these adults had, on average, an eightfold increase in relative abundance in Streptococcus sp., Moraxella catarrhalis, and Haemophilus influenzae, pathobionts known to colonize the children's upper respiratory tract, and a fourfold decrease in Staphylococcus aureus and Staphylococcus lugdunensis.

CONCLUSIONS

Our study showed that, in adults, the presence of S. pneumoniae in the nasopharynx is associated with a shift in the microbiota and dominance of the Streptococcus genus. Furthermore, we observed that smoking habits are associated with an increase in bacterial genera commonly linked to periodontal diseases. Interestingly, our research also revealed that adults who have regular contact with children have a microbiota enriched in pathobionts frequently carried by children. These findings collectively contribute to a deeper understanding of how various factors influence the upper respiratory tract microbiota in adults. Video Abstract.

Risk of bias assessment tool for systematic review and meta-analysis of the gut microbiome

Risk of bias assessment is a critical step of any meta-analysis or systematic review. Given the low sample count of many microbiome studies, especially observational or cohort studies involving human subjects, many microbiome studies have low power. This increases the importance of performing meta-analysis and systematic review for microbiome research in order to enhance the relevance and applicability of microbiome results. This work proposes a method based on the ROBINS-I tool to systematically consider sources of bias in microbiome research seeking to perform meta-analysis or systematic review for microbiome studies.

Zero-Inflated Time Series Clustering Via Ensemble Thick-Pen Transform

This study develops a new clustering method for high-dimensional zero-inflated time series data. The proposed method is based on thick-pen transform (TPT), in which the basic idea is to draw along the data with a pen of a given thickness. Since TPT is a multi-scale visualization technique, it provides some information on the temporal tendency of neighborhood values. We introduce a modified TPT, termed 'ensemble TPT (e-TPT)', to enhance the temporal resolution of zero-inflated time series data that is crucial for clustering them efficiently. Furthermore, this study defines a modified similarity measure for zero-inflated time series data considering e-TPT and proposes an efficient iterative clustering algorithm suitable for the proposed measure. Finally, the effectiveness of the proposed method is demonstrated by simulation experiments and two real datasets: step count data and newly confirmed COVID-19 case data.

Infants' gut microbiome data: A Bayesian Marginal Zero-inflated Negative Binomial regression model for multivariate analyses of count data

The infants' gut microbiome is dynamic in nature. Literature has shown high inter-individual variability of gut microbial composition in the early years of infancy compared to adulthood. Although next-generation sequencing technologies are rapidly evolving, several statistical analysis aspects need to be addressed to capture the variability and dynamic nature of the infants' gut microbiome. In this study, we proposed a Bayesian Marginal Zero-inflated Negative Binomial (BAMZINB) model, addressing complexities associated with zero-inflation and multivariate structure of the infants' gut microbiome data. Here, we simulated 32 scenarios to compare the performance of BAMZINB with glmFit and BhGLM as the two other widely similar methods in the literature in handling zero-inflation, over-dispersion, and multivariate structure of the infants' gut microbiome. Then, we showed the performance of the BAMZINB approach on a real dataset using SKOT cohort (I and II) studies. Our simulation results showed that the BAMZINB model performed as well as those two methods in estimating the average abundance difference and had a better fit for almost all scenarios when the signal and sample size were large. Applying BAMZINB on SKOT cohorts showed remarkable changes in the average absolute abundance of specific bacteria from 9 to 18 months for infants of healthy and obese mothers. In conclusion, we recommend using the BAMZINB approach for infants' gut microbiome data taking zero-inflation and over-dispersion properties into account in multivariate analysis when comparing the average abundance difference.

Variable Selection for Sparse Data with Applications to Vaginal Microbiome and Gene Expression Data

Sparse data with a high portion of zeros arise in various disciplines. Modeling sparse high-dimensional data is a challenging and growing research area. In this paper, we provide statistical methods and tools for analyzing sparse data in a fairly general and complex context. We utilize two real scientific applications as illustrations, including a longitudinal vaginal microbiome data and a high dimensional gene expression data. We recommend zero-inflated model selections and significance tests to identify the time intervals when the pregnant and non-pregnant groups of women are significantly different in terms of Lactobacillus species. We apply the same techniques to select the best 50 genes out of 2426 sparse gene expression data. The classification based on our selected genes achieves 100% prediction accuracy. Furthermore, the first four principal components based on the selected genes can explain as high as 83% of the model variability.

Predicting COVID-19 disease severity from SARS-CoV-2 spike protein sequence by mixed effects machine learning

Epidemiological studies show that COVID-19 variants-of-concern, like Delta and Omicron, pose different risks for severe disease, but they typically lack sequence-level information for the virus. Studies which do obtain viral genome sequences are generally limited in time, location, and population scope. Retrospective meta-analyses require time-consuming data extraction from heterogeneous formats and are limited to publicly available reports. Fortuitously, a subset of GISAID, the global SARS-CoV-2 sequence repository, includes "patient status" metadata that can indicate whether a sequence record is associated with mild or severe disease. While GISAID lacks data on comorbidities relevant to severity, such as obesity and chronic disease, it does include metadata for age and sex to use as additional attributes in modeling. With these caveats, previous efforts have demonstrated that genotype-patient status models can be fit to GISAID data, particularly when country-of-origin is used as an additional feature. But are these models robust and biologically meaningful? This paper shows that, in fact, temporal and geographic biases in sequences submitted to GISAID, as well as the evolving pandemic response, particularly reduction in severe disease due to vaccination, create complex issues for model development and interpretation. This paper poses a potential solution: efficient mixed effects machine learning using GPBoost, treating country as a random effect group. Training and validation using temporally split GISAID data and emerging Omicron variants demonstrates that GPBoost models are more predictive of the impact of spike protein mutations on patient outcomes than fixed effect XGBoost, LightGBM, random forests, and elastic net logistic regression models.

Statistical challenges in longitudinal microbiome data analysis

The microbiome is a complex and dynamic community of microorganisms that co-exist interdependently within an ecosystem, and interact with its host or environment. Longitudinal studies can capture temporal variation within the microbiome to gain mechanistic insights into microbial systems; however, current statistical methods are limited due to the complex and inherent features of the data. We have identified three analytical objectives in longitudinal microbial studies: (1) differential abundance over time and between sample groups, demographic factors or clinical variables of interest; (2) clustering of microorganisms evolving concomitantly across time and (3) network modelling to identify temporal relationships between microorganisms. This review explores the strengths and limitations of current methods to fulfill these objectives, compares different methods in simulation and case studies for objectives (1) and (2), and highlights opportunities for further methodological developments. R tutorials are provided to reproduce the analyses conducted in this review.

Associations of air pollution and greenness with the nasal microbiota of healthy infants: A longitudinal study

BACKGROUND

Air pollution and greenness are associated with short- and long-term respiratory health in children but the underlying mechanisms are only scarcely investigated. The nasal microbiota during the first year of life has been shown to be associated with respiratory tract infections and asthma development. Thus, an interplay between greenness, air pollution and the early nasal microbiota may contribute to short- and long-term respiratory health. We aimed to examine associations between fine particulate matter (PM_2.5), nitrogen dioxide (NO₂) and greenness with the nasal microbiota of healthy infants during the first year of life in a European context with low-to-moderate air pollution levels.

METHODS

Microbiota characterization was performed using 16 S rRNA pyrosequencing of 846 nasal swabs collected fortnightly from 47 healthy infants of the prospective Basel-Bern Infant Lung Development (BILD) cohort. We investigated the association of satellite-based greenness and an 8-day-average exposure to air pollution (PM_2.5, NO₂) with the nasal microbiota during the first year of life. Exposures were individually estimated with novel spatial-temporal models incorporating satellite data. Generalized additive mixed models adjusted for known confounders and considering the autoregressive correlation structure of the data were used for analysis.

RESULTS

Mean (SD) PM_2.5 level was 17.1 (3.8 μg/m³) and mean (SD) NO₂ level was 19.7 (7.9 μg/m³). Increased PM_2.5 and increased NO₂ were associated with reduced within-subject Ružička dissimilarity (PM_2.5: per 1 μg/m³ -0.004, 95% CI -0.008, -0.001; NO₂: per 1 μg/m³ -0.004, 95% CI -0.007, -0.001). Whole microbial community comparison with nonmetric multidimensional scaling revealed distinct microbiota profiles for different PM_2.5 exposure levels. Increased NO₂ was additionally associated with reduced abundance of Corynebacteriaceae (per 1 μg/m³: -0.027, 95% CI -0.053, -0.001). No associations were found between greenness and the nasal microbiota.

CONCLUSION

Air pollution was associated with Ružička dissimilarity and relative abundance of Corynebacteriaceae. This suggests that even low-to-moderate exposure to air pollution may impact the nasal microbiota during the first year of life. Our results will be useful for future studies assessing the clinical relevance of air-pollution-induced alterations of the nasal microbiota with subsequent respiratory disease development.

MZINBVA: variational approximation for multilevel zero-inflated negative-binomial models for association analysis in microbiome surveys

As our understanding of the microbiome has expanded, so has the recognition of its critical role in human health and disease, thereby emphasizing the importance of testing whether microbes are associated with environmental factors or clinical outcomes. However, many of the fundamental challenges that concern microbiome surveys arise from statistical and experimental design issues, such as the sparse and overdispersed nature of microbiome count data and the complex correlation structure among samples. For example, in the human microbiome project (HMP) dataset, the repeated observations across time points (level 1) are nested within body sites (level 2), which are further nested within subjects (level 3). Therefore, there is a great need for the development of specialized and sophisticated statistical tests. In this paper, we propose multilevel zero-inflated negative-binomial models for association analysis in microbiome surveys. We develop a variational approximation method for maximum likelihood estimation and inference. It uses optimization, rather than sampling, to approximate the log-likelihood and compute parameter estimates, provides a robust estimate of the covariance of parameter estimates and constructs a Wald-type test statistic for association testing. We evaluate and demonstrate the performance of our method using extensive simulation studies and an application to the HMP dataset. We have developed an R package MZINBVA to implement the proposed method, which is available from the GitHub repository https://github.com/liudoubletian/MZINBVA.