Biology subjects

Gim, J.

Publications and source records attributed to Gim, J..

6 recordsLinked to original sources

bioRxiv2025

Microstructural Spine Alterations Increase Neuronal Excitability in Focal Cortical Dysplasia

Focal cortical dysplasia(FCD) is a leading cause of drug-resistant epilepsy, traditionally attributed to impaired inhibition. However, recent ultrastructural evidence suggests excitatory synaptic alterations may also contribute. To investigate this, we performed computational modeling of human pyramidal neurons in NEURON, incorporating electron microscopy-derived dendritic spine morphologies. Baseline and FCD-like models were constructed with differences in spine density and morphology while maintaining comparable net synaptic input. Simulations revealed that reduced spine density increased input resistance and amplified excitatory postsynaptic potentials, enhancing dendritic-to-somatic transmission. Spine neck geometry, but not head size, critically shaped voltage compartmentalization. The altered model required markedly fewer synchronous inputs to reach spiking threshold and exhibited higher firing rates under Poisson-distributed synaptic activity, particularly under sparse input conditions. These results demonstrate that excitatory microstructural changes - reduced spine density and weakened neck compartmentalization - can elevate neuronal excitability, in addition to the inhibitory deficit which is the biggest known effect. These findings highlight excitatory spine remodeling as another key driver of epileptogenesis. We speculate that extending this analysis will be essential to fully understand circuit-level hyperexcitability and to identify new therapeutic targets.

neuroscience↗

bioRxiv2024

Unlocking the Potential of Low Quality Total RNA-seq Data: A Stepwise Mapping Approach for Improved Quantitative Analyses

High-throughput sequencing assays face persistent challenges when analyzing low-quality RNAs, often assessed by the RNA integrity number (RIN). Current preprocessing methods and pipelines designed for mRNA-seq presume high-quality RNAs, overlooking the nuanced complexities arising from degraded transcripts in low-quality samples. This study questions the applicability of standard analysis pipelines, especially when sequencing low-quality total RNAs, which are sometimes the sole recourse for specific inquiries. To address this, we conducted a comprehensive investigation using large sequencing reads obtained from blood biospecimens with varying RINs. Introducing a novel mapping approach, termed stepwise mapping, our systematic comparative analyses propose an optimal practice for total RNA-seq data analysis. Contrary to conventional mapping procedures, the stepwise mapping approach unveils additional transcriptome information, crucial for stable differential expression analysis, even with total RNA-seq data from specimens with relatively low RINs. Our method proves particularly valuable when analyzing limited specimens with low RNA quality.

bioinformatics↗

bioRxiv2023

A cerebellar disinhibitory circuit supports synaptic plasticity

How does the cerebellum learn how to control motion? The cerebellar motor learning critically depends on the long-term depression of the synapses between granule cells and Purkinje cells, which encode motor commands and inhibitory modifications to motor outputs, respectively, for simultaneous granule cell inputs and climbing fibre inputs, the latter of which encode the error signals1-3. However, recent studies have revealed that inhibitory inputs to Purkinje cells may disrupt long-term depression4-8, and it is not clear how long-term depression can occur without disruption. In search of a clue, we investigated the synaptic connectivity among the neurons reconstructed from serial electron microscopy images of the cerebellar molecular layer9,10. We discovered synapses between climbing fibres and a subset of inhibitory interneurons, which synapse onto the remaining interneurons, which in turn synapse onto Purkinje cells. Such connectivity redefines the interneuron types, which have been defined morphologically or molecularly11-13. Together with climbing fibres to Purkinje cell connections, those cell types form a feedforward disinhibitory circuit14. We argued that this circuit secures long-term depression by suppressing inhibition whenever climbing fibre input is provided and long-term depression needs to occur15, and we validated the hypothesis through a computational model. This finding implies a general principle of circuit mechanism in which disinhibition supports synaptic plasticity16,17.

neuroscience↗

bioRxiv2023

GraphMHC: neoantigen prediction model applying the graph neural network to molecular structure

Neoantigens are biomarkers that can predict the prognosis associated with immune checkpoint inhibition by estimating the binding potential of candidate peptides to somatic mutation and major histocompatibility complex (MHC) proteins. Although deep neural networks have been primarily used for these prediction models, it is difficult to consider the models reported thus far as accurately representing the interactions between biomolecules. In this study, we propose the GraphMHC model, which utilizes a graph neural network model through molecular structure to simulate the binding between MHC proteins and peptide sequences. Amino acid sequences sourced from the immune epitope database (IEDB) undergo conversion into molecular structures. Subsequently, atomic intrinsic informations and inter-atomic connections are extracted and structured as a graph representation. Bindings are classified by feeding them into the GraphMHC network, comprising stacked graph attention and convolution layers. The prediction results from the test set using the GraphMHC model showed a high performance with an area under the receiver operating characteristic curve of 92.2% (91.9-92.5%), surpassing the baseline model. Moreover, by applying the GraphMHC model to melanoma patient data from the Cancer Genome Atlas project, we found a borderline difference in overall survival and a significant difference in stromal score between the high and low neoantigen load groups. This distinction was not present in the baseline model. This study presents the first feature-intrinsic method based on biochemical molecular structure for modeling the binding between MHC protein sequences and neoantigen candidate peptide sequences. The model can provide highly accurate suitability information for cancer patients who want to apply immune checkpoint inhibitors. Author summary

bioinformatics↗

bioRxiv2022

DeepParcellation: a novel deep learning method for robust brain magnetic resonance imaging parcellation in older East Asians

Accurate parcellation of cortical regions is crucial for distinguishing morphometric changes in aged brains, particularly in degenerative brain diseases. Normal aging and neurodegeneration precipitate brain structural changes, leading to distinct tissue contrast and shape in people aged > 60 years. Manual parcellation by trained radiologists can yield a highly accurate outline of the brain; however, analyzing large datasets is laborious and expensive. Alternatively, newly-developed computational models can quickly and accurately conduct brain parcellation, although thus far only for the brains of Caucasian individuals. DeepParcellation, our novel deep learning model for 3D magnetic resonance imaging (MRI) parcellation, was trained on 5,035 brains of older East Asians (Gwangju Alzheimers & Related Dementia) and 2,535 brains of Caucasians. We trained full 3D models for N-way individual regions of interest using memory reduction techniques. Our method showed the highest similarity and robust reliability among age-ethnicity groups, especially when parcellating the brains of older East Asians.

neuroscience↗

bioRxiv2018

BALLI: Bartlett-Adjusted Likelihood-based LInear Model Approach for Identifying Differentially Expressed Gene with RNA-seq Data

MotivationTranscriptomic profiles can improve our understanding of the phenotypic molecular basis of biological research, and many statistical methods have been proposed to identify differentially expressed genes under two or more conditions with RNA-seq data. However, statistical analyses with RNA-seq data often suffer from small sample sizes, and global variance estimates of RNA expression levels have been utilized as prior distributions for gene-specific variance estimates, making it difficult to generalize the methods to more complicated settings. We herein proposed a Bartlett-Adjusted Likelihood based LInear mixed model approach (BALLI) to analyze more complicated RNA-seq data. The proposed method estimates the technical and biological variances with a linear mixed effect model, with and without adjusting small sample bias using Bartletts corrections.\n\nResultsWe conducted extensive simulations to compare the performance of BALLI with those of existing approaches (edgeR, DESeq2, and voom). Results from the simulation studies showed that BALLI correctly controlled the type-1 error rates at the various nominal significance levels, and produced better statistical power and precision estimates than those of other competing methods in various scenarios. Furthermore, BALLI was robust to variation of library size. It was also successfully applied to Holstein milk yield data, illustrating its practical value.\n\nAvailability and ImplementationBALLI is implemented as R package and freely available at http://healthstat.snu.ac.kr/software/balli/.\n\nContactwon1@snu.ac.kr\n\nSupplementary InformationSupplementary data are available at Bioinformatics online

bioinformatics↗