Biology subjects

Randolph, T. W.

Publications and source records attributed to Randolph, T. W..

6 recordsLinked to original sources

bioRxiv2026

Mediation Analysis with Compositional Exposures

Understanding the causal pathways linking the gut microbiome to downstream biomarkers and clinical outcomes is central to microbiome research. When the microbiome acts as the exposure, the high dimensionality and compositional structure of the data introduce unique methodological challenges for mediation analysis. To address these challenges, we propose a latent variable mediation framework that captures variation in microbial composition through a microbial balance, defined as the log-ratio between two unknown subsets of taxa. This balance serves as a latent scalar exposure and simplifies the estimation of the overall indirect effect at the community level, while simultaneously identifying specific taxa that contribute to the overall direct and indirect effects. We describe the model's estimation and inference, and illustrate the method using data from the Multiethnic Cohort--Adiposity Phenotype Study, where we examine the mediation pathway from the gut microbiome through lipopolysaccharide-binding protein, a marker of metabolic endotoxemia, to percent liver fat.

bioinformatics↗

bioRxiv2026

Atomic layer deposition for core-shell microparticle vaccines enabling programmable antigen delivery to lymph nodes enhance humoral immune responses

Technologies that simplify complex dosing regimens as single-shot immunizations may be important for vaccines against difficult to neutralize pathogens. Here we characterized delivery mechanisms of core-shell microparticles comprising clinically-relevant HIV immunogen formulations spray-dried to form solid spherical microparticles and subsequently coated with a nanoscopic alumina shell using atomic layer deposition (ALD). ALD particles exhibited a time delay in antigen release programmed by the alumina shell followed by prolonged antigen release, which steadily accumulated in antigen-presenting cells at the injection site and draining lymph nodes and accumulated on follicular dendritic cells in B cell follicles. ALD vaccines elicited continuous expansion of antigen-specific germinal center B cells over 8 weeks, serum antibody responses with >10-fold slower antigen-binding off-rates, and 2-fold more long-lived plasma cells compared to traditional bolus vaccination with potent adjuvants. Single administration ALD technology thus promotes key events in the primary immune response important for vaccines against HIV and other challenging pathogens.

immunology↗

bioRxiv2026

Gut Microbiota Mediates the Association between Diet Quality and Ectopic Adiposity: The Multiethnic Cohort Adiposity Phenotype Study

BackgroundHigher-quality diets have been associated with lower levels of ectopic fat deposited in the viscera and liver, which is hypothesized to be mediated in part by the gut microbiota. ObjectivesWe tested this hypothesis in a multi-ethnic imaging study using global (microbiome-wide) testing as well as a high-dimensional multiple-mediators regression framework to identify bacterial genera in the human gut that mediate the association between diet quality and ectopic adiposity. MethodsWe analyzed the cross-sectional data of 1,400 older adults (age 60-77) from five racial/ethnic groups in the Multiethnic Cohort Adiposity Phenotype Study (2013-2016). Overall diet quality was defined by adherence to the MIND diet. The relative abundance of 151 bacterial genera was quantified from 16S rRNA gene sequencing of the stool samples. Visceral fat, liver fat, and the presence of MASLD (metabolic dysfunction-associated steatotic liver disease) were determined based on magnetic resonance imaging (MRI). We used high-dimensional mediation analysis (HDMA) to estimate gut microbial mediation in the linear regression of visceral fat or liver fat, or in logistic regression of MASLD, on the MIND adherence score, adjusted for potential confounders. ResultsHigher diet quality was associated with lower ectopic adiposity: 12% less visceral fat area, 23% less liver fat, and a 49% less likelihood of having MASLD, comparing the highest to the lowest quartile of the MIND score. Using a distance-based global test, we confirmed overall significant microbial mediation of the inverse diet-ectopic fat association. From HDMA, four bacterial genera were identified as mediating the protective association with visceral fat, with the largest mediation conferred by Lachnospiraceae UCG.001 (12.2%). Two genera (Lachnoclostridium, Weissella) were shown to mediate the MIND association with both liver fat and MASLD. In particular, Lachnoclostridium mediated 13.6% of the liver fat association and 10.8% of the MASLD association, and Lachnospiraceae UCG.001 additionally mediated 12.1% of the liver fat association. ConclusionsOur results support the hypothesis that the gut microbiota contributes to conveying the effect of diet quality on preferred body fat distribution, e.g., involving bacteria that are known to produce short-chain fatty acids (Lachnospiraceae) or secondary bile acids (Lachnoclostridium).

microbiology↗

bioRxiv2025

Single administration of mosaic-8b RBD-nanoparticle vaccine prepared with atomic layer deposition technology elicits broadly neutralizing anti-sarbecovirus responses

Atomic layer deposition (ALD), a new vaccine technology, permits multiple dosing with a single administration by pulsatile release of one or more immunogens. We evaluated ALD delivery of mosaic-8b [60-mer nanoparticles presenting 8 different SARS-like betacoronavirus (sarbecovirus) receptor-binding domains (RBDs)] that elicits broadly cross-reactive antibodies and protects against mismatched sarbecoviruses not represented by RBDs on mosaic-8b. Compared with conventional prime-boost immunizations, ALD-delivered mosaic-8b RBD-nanoparticles elicited antibodies in both naive and pre-vaccinated mice with improved mismatched binding and neutralization. Results of RBD epitope mapping of serum antibodies from ALD-delivered mosaic-8b were consistent with broader coverage of RBD epitopes compared to conventional immunizations, and systems serology revealed distinct IgG subclass and Fc{gamma}R-binding IgG distributions. These results suggest that ALD is a promising technology for use with mosaic-8b RBD-nanoparticle vaccines to protect against future sarbecovirus spillovers and support applications for ALD vaccine delivery to elicit cross-reactive antibodies against rapidly mutating or diverse pathogens.

immunology↗

bioRxiv2018

Connectivity-Informed Adaptive Regularization for Generalized Outcomes

One of the challenging problems in the brain imaging research is a principled incorporation of information from different imaging modalities in association studies. Frequently, data from each modality is analyzed separately using, for instance, dimensionality reduction techniques, which result in a loss of mutual information. We propose a novel regularization method, griPEER (generalized ridgified Partially Empirical Eigenvectors for Regression) to estimate the association between the brain structure features and a scalar outcome within the generalized linear regression framework. griPEER provides a principled approach to use external information from the structural brain connectivity to improve the regression coefficient estimation. Our proposal incorporates a penalty term, derived from the structural connectivity Laplacian matrix, in the penalized generalized linear regression. We address both theoretical and computational issues and show that our method is robust to the incomplete information about the structural brain connectivity. We also provide a significance testing procedure for performing inference on the estimated coefficients in this model. griPEER is evaluated in extensive simulation studies and it is applied in classification of the HIV+ and HIV- individuals.

neuroscience↗

bioRxiv2017

Brain connectivity-informed regularization methods for regression

A challenging problem arising in brain imaging research is principled incorporation of information from different imaging modalities. Frequently each modality is analyzed separately using, for instance, dimensionality reduction techniques which result in a loss of mutual information. We propose a novel regularization method to estimate the association between the brain structure features and a scalar outcome within the linear regression framework. Our regularization technique provides a principled approach to utilizing external information arising from the structural brain connectivity to inform the estimation of the regression coefficients. Our proposal extends the classical Tikhonov regularization framework by defining a penalty term based on the structural connectivity-derived Laplacian matrix. In the work presented, we address both theoretical and computational issues. The approach is illustrated using simulated data and compared with other penalized regression methods. Finally, we apply our regularization method to study the associations between the alcoholism phenotypes and brain cortical thickness using a diffusion tensor imaging (DTI) derived measure of structural connectivity.

neuroscience↗