Biology subjects

Hyun, S.

Publications and source records attributed to Hyun, S..

4 recordsLinked to original sources

bioRxiv2026

Responses to unannounced rule changes in adults with developmental disabilities: A community-based pilot study with consumer-grade EEG

Unannounced changes in rules or environments often precede challenging behavior and limit community participation for adults with developmental disabilities, yet their responses to change have rarely been quantified directly. We recruited adults with developmental disabilities, including those with intellectual disability, at a community day-service center. We administered a probabilistic reversal learning task and recorded resting-state electroencephalography (EEG) via a 14-channel wireless EEG system. Among the 12 adults who performed the task, total errors were comparable across participants (range 31-57), whereas a switch bias index, contrasting perseverative errors (persisting with the previously correct option) with regressive errors (failing to maintain the newly correct option), ranged from -0.96 to +0.91. In the 9 participants included in resting-state EEG analyses, a flatter aperiodic spectrum over the frontal cluster (preregistered, but selected as the focus post hoc) was associated with fewer switches after negative feedback ({rho} = .80) and more perseverative errors ({rho} = -.72). The preregistered parieto-occipital cluster showed the same direction ({rho} = .69, 95% confidence interval [-.11, .97], n = 8), which was opposite to the preregistered prediction. Frontal alpha peak amplitude was associated with accuracy ({rho} = .77), and the between-run reliability of resting-state measures was .82-.99. These findings provide preliminary evidence that EEG recorded in community settings can capture individual differences in how people respond to change. Field measurement that includes people with intellectual disability may help extend the evidence base for the Sustainable Development Goals of good health and well-being (SDG 3) and reduced inequalities (SDG 10).

neuroscience↗

bioRxiv2021

A flexible Bayesian approach to estimating size-structured matrix population models

The rates of cell growth, division, and carbon loss of microbial populations are key parameters for understanding how organisms interact with their environment and how they contribute to the carbon cycle. However, the invasive nature of current analytical methods has hindered efforts to reliably quantify these parameters. In recent years, size-structured matrix population models (MPMs) have gained popularity for estimating rate parameters of microbial populations by mechanistically describing changes in microbial cell size distributions over time. And yet, the construction, analysis, and biological interpretation of these models are underdeveloped, as current implementations do not adequately constrain or assess the biological feasibility of parameter values, leading to inference which may provide a good fit to observed size distributions but does not necessarily reflect realistic physiological dynamics. Here we present a flexible Bayesian extension of size-structured MPMs for testing underlying assumptions describing the dynamics of a marine phytoplankton population over the day-night cycle. Our Bayesian framework takes prior scientific knowledge into account and generates biologically interpretable results. Using data from an exponentially growing laboratory culture of the cyanobacterium Prochlorococcus, we herein demonstrate the performance improvements of our approach over current models and isolate previously ignored biological processes, such as respiratory and exudative carbon losses, as critical parameters for the modeling of microbial population dynamics. The results demonstrate that this modeling framework can provide deeper insights into microbial population dynamics provided by flow-cytometry time-series data. Author summaryIdentifying the growth and population dynamics of marine microorganisms in their natural habitat is crucial to understanding the flow of carbon in the oceans but remains a grand challenge due to the invasive nature of current measurement methods. As time-series observations of population size structure have become more commonplace in aquatic environments, matrix population models (MPMs), which aim to mechanistically describe the change in size structure of these populations over time, have gained in popularity over the last decade. However, the underlying assumptions and behavior of MPMs have not been adequately scrutinized, and parameter values are difficult to interpret biologically, leading to inference that may not reflect plausible physiological dynamics. Here, we develop a Bayesian extension of the MPM framework to examine biological assumptions, improve interpretability of model output, and account for additional biological processes. We evaluated the performance of our models on a publicly available dataset of laboratory experiment time-series measurements of the cyanobacterium Prochlorococcus, Earths most abundant photosynthetic organisms, demonstrated the performance improvements of our approach over current models, and isolated previously ignored respiratory and exudative carbon losses as critical parameters for the modeling of microbial population dynamics.

microbiology↗

bioRxiv2020

Deep Metric Learning Network using Proxies for Chromosome Classification and Retrieval in Karyotyping Test

In karyotyping, the classification of chromosomes is a tedious, complicated, and time-consuming process. It requires extremely careful analysis of chromosomes by well-trained cytogeneticists. To assist cytogeneticists in karyotyping, we introduce Proxy-ResNeXt-CBAM which is a metric learning based network using proxies with a convolutional block attention module (CBAM) designed for chromosome classification. RexNeXt-50 is used as a backbone network. To apply metric learning, the fully connected linear layer of the backbone network (ResNeXt-50) is removed and is replaced with CBAM. The similarity between embeddings, which are the outputs of the metric learning network, and proxies are measured for network training. Proxy-ResNeXt-CBAM is validated on a public chromosome image dataset, and it achieves an accuracy of 95.86%, a precision of 95.87%, a recall of 95.9%, and an F-1 score of 95.79%. Proxy-ResNeXt-CBAM which is the metric learning network using proxies outperforms the baseline networks. In addition, the results of our embedding analysis demonstrate the effectiveness of using proxies in metric learning for optimizing deep convolutional neural networks. As the embedding analysis results show, Proxy-ResNeXt-CBAM obtains a 94.78% Recall@1 in image retrieval, and the embeddings of each chromosome are well clustered according to their similarity.

bioengineering↗

bioRxiv2020

An alternative miRISC targeting a coding mutation site in FOXL2 links to granulosa cell tumor

Recent evidence suggests that animal microRNAs (miRNAs) can target coding sequences (CDSs); however, the pathophysiological importance of such targeting remains unknown. Here, we show that a somatic heterozygous missense mutation (c.402C>G; p.C134W) in FOXL2, a feature shared by virtually all adult-type granulosa cell tumors (AGCTs), introduces a target site for miR-1236, which induces haploinsufficiency of the tumor-suppressor FOXL2. This miR-1236-mediated selective degradation of the variant FOXL2 mRNA is preferentially conducted by a distinct miRNA-loaded RNA-induced silencing complex (miRISC) directed by the Argonaute3 (AGO3) and DHX9 proteins. In both patients and mouse model of AGCT, the inversely regulated variant FOXL2 abundance with the miR-1236 levels was highly correlated with malignant features of AGCT. Our study provides a molecular basis for understanding the conserved FOXL2 CDS mutation-mediated etiology of AGCT, revealing the existence of a previously unidentified mechanism of miRNA-targeting disease-associated mutations in the CDS by forming a non-canonical miRISC.

cancer biology↗