Biology subjects

Huynh, A. V.

Publications and source records attributed to Huynh, A. V..

4 recordsLinked to original sources

bioRxiv2026

StrataBionn: a neural network supervised classification method for microbial communities

The classification of microbial communities into discrete states or "community state types" (CSTs) is fundamental to understanding host-microbiome interactions and their clinical implications. Traditional methods, such as the nearest-neighbor approaches, often struggle with the inherent noise, high dimensionality, and non-linear signatures of taxonomic profiles. We present a novel supervised framework for microbial community classification, leveraging an Artificial Neural Network (ANN) architecture implemented in a new tool we named StrataBionn. We rigorously evaluated our approach using large-scale vaginal microbiome datasets, directly benchmarking performance against VALENCIA and a Random Forest (RF) classifier. To demonstrate the versatility of our models, we further extended the framework to oral microbiome classification, assessing its stability across diverse anatomical sites. Our supervised models consistently outperformed the nearest-neighbor approach across all evaluated datasets. In the vaginal microbiome, our method achieved an 11.6% to 13.3% increase in performance across all primary metrics, including precision, recall, accuracy, and F1-score. Furthermore, we demonstrate that this performance advantage is maintained in the oral microbiome, highlighting the generalizability of our neural network and ensemble strategies to various microbial ecosystems without the need for niche-specific algorithmic adjustments. By capturing complex feature dependencies that distance-based methods overlook, our approach provides a more robust and accurate census of microbial community structures. StrataBionns ability to learn classification schemes for any microbiome with high accuracy and explainability, through the use of provided utilities to visualize feature-space classification boundaries and perform perturbation analysis on trained classifiers, makes it ideal for broad application in microecology research. This framework offers a scalable, high-performance alternative for microbiome researchers, facilitating more precise clinical stratification and biological insights across hosts body sites.

genomics↗

bioRxiv2024

Dominance reversal protects large-effect resistance polymorphisms in temporally varying environments

Large-effect functional genetic variation is commonly found in natural populations, even though natural selection should erode such variants. Theory suggests that under fluctuating selective pressures, beneficial reversal of dominance - where alleles are dominant when beneficial and recessive when deleterious - can protect these loci from selection, allowing them to persist. However, empirical evidence for this mechanism remains elusive because testing requires direct measurements of selection and dominance in natural conditions. Here, we show that insecticide-resistant alleles at the Ace locus in Drosophila melanogaster persist worldwide at intermediate frequencies and exhibit beneficial reversal of dominance. By combining laboratory and large-scale field mesocosm experiments with insecticide manipulation, and mathematical modeling, we show that the benefits of the resistant Ace alleles are dominant while their fitness costs recessive. We further show that fluctuating insecticide selection generates chromosome-scale genomic perturbations at sites linked to the resistant Ace alleles, revealing broader genomic consequences of this mechanism. Overall, our results suggest that beneficial reversal of dominance contributes to the maintenance of functional genetic variation and impacts patterns of genomic diversity via linked fluctuating selection.

evolutionary biology↗

bioRxiv2022

Small molecule-mediated insulin hypersecretion induces transient unfolded protein response and loss of function in pancreatic islet beta cells

Pancreatic islet beta cells require a fine-tuned ER stress response for normal function; abnormal ER stress contributes to diabetes pathogenesis. Here, we identified a small molecule, SW016789, with time-dependent effects on beta cell ER stress and function. Acute treatment with SW016789 potentiated nutrient-induced calcium influx and insulin secretion, while chronic exposure to SW016789 transiently induced ER stress and shut down secretory function in a reversible manner. Distinct from the effects of thapsigargin, SW016789 did not affect beta cell viability or apoptosis, potentially due to a rapid induction of adaptive genes, weak signaling through the eIF2 kinase PERK, and lack of oxidative stress gene Txnip induction. We determined that SW016789 acted upstream of voltage-dependent calcium channels (VDCCs) and potentiated nutrient- but not KCl-stimulated calcium influx. Measurements of metabolomics, oxygen consumption rate, and G protein-coupled receptor signaling did not explain the potentiating effects of SW016789. In chemical co-treatment experiments we discovered synergy between SW016789 and activators of protein kinase C (PKC) and VDCCs, suggesting involvement of these pathways in the mechanism of action. Finally, chronically elevated calcium influx was required for the inhibitory impact of SW016789, as blockade of VDCCs protected human islets and MIN6 beta cells from hypersecretion-induced dysfunction. We conclude that beta cells undergoing this type of pharmacological hypersecretion have the capacity to suppress their function to mitigate ER stress and avoid apoptosis. These results have the potential to uncover beta cell ER stress mitigation factors and add support to beta cell rest strategies to preserve function.

cell biology↗

bioRxiv2017

Hybrid Songbirds are Deficient in Learning and Memory

Identifying the phenotypes underlying postzygotic reproductive isolation is crucial for fully understanding the evolution and maintenance of species. One potential postzygotic isolating barrier that has not yet been examined is learning and memory ability in hybrids. Learning and memory are important fitness-related traits, especially in scatter-hoarding species, where accurate retrieval of hoarded food is vital for winter survival. Here, we test the hypothesis that learning and memory ability can act as a postzygotic isolating barrier by comparing these traits among two scatter-hoarding songbird species, black-capped (Poecile atricapillus), Carolina chickadees (Poecile carolinensis), and their naturally occurring hybrids. In an outdoor aviary setting, we find that hybrid chickadees perform significantly worse on an associative learning spatial task and are worse at solving a novel problem compared to both parental species. Deficiencies in learning and memory abilities could therefore contribute to postzygotic reproductive isolation between chickadee species. Given the importance of learning and memory for fitness, our results suggest that these traits may play an important, but as yet overlooked, role in postzygotic reproductive isolation.

evolutionary biology↗