A5101931124- Genomics and Phylogenetic Studies
- Biomedical Text Mining and Ontologies
- RNA and protein synthesis mechanisms
- Bioinformatics and Genomic Networks
- Microbial Metabolic Engineering and Bioproduction
- Genomics and Chromatin Dynamics
- Genetics, Bioinformatics, and Biomedical Research
- Cutaneous Melanoma Detection and Management
- RNA modifications and cancer
- AI in cancer detection
- Machine Learning in Bioinformatics
- Topic Modeling
- Genomics and Rare Diseases
- Gene Regulatory Network Analysis
- Biomedical and Engineering Education
- Epigenetics and DNA Methylation
- Natural Language Processing Techniques
- Cancer-related molecular mechanisms research
- Digital Media Forensic Detection
- Algorithms and Data Compression
- COVID-19 epidemiological studies
- Radiation Dose and Imaging
- Bacillus and Francisella bacterial research
- Ferroptosis and cancer prognosis
- Data-Driven Disease Surveillance
Ewha Womans University
2015-2025
National University College
2012
National Cancer Center
2012
Ewha Womans University Medical Center
2008-2011
Sejong University
2000
Globally, breast cancer (BC) is considered a major cause of death among women. Therefore, researchers have used various machine and deep learning-based methods for its early accurate detection using X-ray, MRI, mammography image modalities. However, the learning model requires domain experts to select an optimal feature, obtains limited accuracy, has high false positive rate due handcrafting features extraction. The overcomes these limitations, but models require large amounts training data...
Abstract Summary: FX is an RNA-Seq analysis tool, which runs in parallel on cloud computing infrastructure, for the estimation of gene expression levels and genomic variant calling. In mapping short reads, uses a transcriptome-based reference primarily, generated from ~160 000 mRNA sequences RefSeq, UCSC Ensembl databases. This approach reduces misalignment reads originating splicing junctions. Unmapped not aligned known transcripts are then mapped human genome reference. allows data...
The early and accurate detection of skin cancer can reduce mortality rates improve patient outcomes, but requires advanced diagnostics. integration artificial intelligence (AI) into healthcare enables the precise timely cancer. However, significant challenges remain including difficulty in differentiating visually similar conditions limitations diverse, representative datasets. In this study, we proposed DCAN-Net, a novel deep-learning framework designed for model leverages an efficient...
Skin cancer, particularly melanoma, is a severe health threat that necessitates early detection for effective treatment. This research introduces skin lesion classification system harnesses the capabilities of three advanced deep learning models: VGG16, Inception-V3, and ResNet-50. By integrating these models into an ensemble, leverages their individual strengths to improve accuracy robustness. Every model in ensemble brings its distinctive contribution, having undergone pre-training on...
This study addresses challenges in skin cancer detection, particularly issues like class imbalance and the varied appearance of lesions, which complicate segmentation classification tasks. The research employs deep learning ensemble models for both (using U-Net, SegNet, DeepLabV3) VGG16, ResNet-50, Inception-V3). ISIC dataset is balanced through oversampling classification, preprocessing techniques such as data augmentation post-processing are applied to increase robustness. model...
Treating genomes just as languages raises the possibility of producing concise generalizations about information in biological sequences.Grammars used this way would constitute a model underlying processes or structures, and that grammars may, fact, serve an appropriate tool for theory formation.The increasing number sequences have been yielded further highlights growing need developing grammatical systems bioinformatics.The intent review is therefore to list some bibliographic references...
Genomics & Informatics (NLM title abbreviation: Inform) is the official journal of Korea Genome Organization. Text corpus for this annotated with various levels linguistic information would be a valuable resource as process extraction requires syntactic, semantic, and higher natural language processing. In study, we publish our new called GNI Corpus version 1.0, extracted from full texts Informatics, NLTK (Natural Language ToolKit)-based text mining script. The preliminary could used...
Epigenetic computational analyses based on Markov chains can integrate dependencies between regions in the genome that are directly adjacent.In this paper, BED files of fifteen chromatin states Broad Histone Track ENCODE project parsed, and comparative nucleotide frequencies regional blocks thoroughly analyzed to detect property them.We perform various tests examine embedded a frequency domain by checking for presence states.We apply these each region states.The results our simulation...
As viruses evolve rapidly, variations in their DNA may arise due to environmental factors. This study examines the classification of COVID-19 sequences based on country origin and analyzes primary correlation with country’s international travel policy. Focusing from nine ASEAN countries, we conducted a two-class distinguish individual countries mixed others. The were initially dissected into 200 base pair units, deep-learning method was employed construct model. Our results showcase capacity...
Integrating various pathway data collections to create new biological knowledge is a challenge, for which novel computational tools play key role.For this purpose, we developed the Java-based conversion modules KGML2SBML and KGML2BioPAX translate KGML (KEGG Markup Language) into couple of common exchange formats: SBML (Systems Biology BioPAX (Biological Pathway Exchange).We hope that our work will be beneficial other Java developers when they extend their bioinformatics system SBML-or...
The static approach of representing metabolic pathway diagrams offers no flexibility.Thus, many systems adopt automatic graph layout techniques to visualize the topological architecture pathways.There are weaknesses, however, because automatically drawn figures generally difficult understand.The problem becomes even more serious when we attempt all information in a single, big picture, which usually results confusing diagram.To provide partial solution this thorny issue, propose J2dpathway,...
The basic graph layout technique, one of many visualization techniques, deals with the problem positioning vertices in a way to maximize some measure desirability graph.The technique is becoming critically important for further development field systems biology.However, applying appropriate automatic techniques genomic scale flow metabolism requires an understanding characteristics and patterns duplicate shared vertices, which crucial bioinformatics software developers.In this paper, we...
The characteristics of metabolic pathways make them particularly amenable to layered graph drawing methods.This paper presents a visual Java-based tool for and annotating biological in twoand-a-half dimensions (2.5D) as an alternative threedimensional (3D) visualizations.Such visualization allows user display different groups clustered nodes, parallel planes, see detailed view group objects focus its place the context whole system.This is extended version J2dPathway.
Recently, the necessity of using low-dose CT imaging with reduced noise has come to forefront due risks involved in radiation. In order acquire a high-resolution image from low-resolution which produces relatively small amount radiation, various algorithms including deep learning-based methods have been proposed. However, current techniques shown limited performance, especially regard losing fine details and blurring high-frequency edges. To enhance previously suggested 2D patch-based...
A large proportion of human noncoding DNA had been known to have no biological function. However, unprecedented technical advances started convert unannotated into highly annotated functional regions. In this paper, the frequency n-grams regional sequences from fifteen chromatin states Broad Histone Track are thoroughly analyzed, applying language modelling n-grams. It has shown that a few particular found in abundance one state but occurring very rarely other states, thereby serving as...
Grammatical inference methods are expected to find grammatical structures hidden in biological sequences. One hopes that studies of grammar serve as an appropriate tool for theory formation. Thus, we have developed JSequitur automatically generating the structure sequences framework string compression algorithms. Our original motivation was any traits several cancer genes can be detected by Through this research, could not meaningful unique yet, but observe some interesting regards...
To allow for a quick conversion of the proprietary sequence data from various sequencing platforms, format toolkits are required that can be easily integrated into workflow systems.In this respect, tool, as well quality tool would minimum requirements to integrate reads different platforms.We have developed Pyrus NGS Sequencing Format Converter, simple software toolkit which allows convert three kinds Next Generation reads, commonly used fasta or fastq formats.The converter modules all...
The non-coding DNA in eukaryotic genomes encodes a language that programs chromatin accessibility, transcription factor binding, and various other activities. objective of this study was to determine the effect primary sequence on epigenomic landscape across 200-base pair genomic units by integrating 127 publicly available ChromHMM BED files from Roadmap Genomics project. Nucleotide frequency profiles annotations stratified variability were analyzed integrative hidden Markov models built...