JOURNALS
2025
47.
Chaewon Kim; Sunyong Yoo
BibTeX | Tags:
@unpublished{nokeyc,
title = {Association between Nutrient Intakes and Osteoporosis in the United States and South Korea: A Population-based Study},
author = {Chaewon Kim and Sunyong Yoo },
year = {2025},
date = {2025-12-31},
urldate = {2024-12-28},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
46.
Yeabean Na; Junho Kim; Myung-Gyun Kang; Sunyong Yoo
BibTeX | Tags:
@unpublished{nokey,
title = {A Multimodal Deep Learning Approach for Predicting Drug Metabolism According to CYP2D6 Genetic Variation},
author = {Yeabean Na and Junho Kim and Myung-Gyun Kang and Sunyong Yoo},
year = {2025},
date = {2025-12-31},
urldate = {2024-12-30},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
45.
Md Sanzid Bin Hossain; Hwan Choi; Zhishan Guo; Sunyong Yoo; Min-Keun Song; Hyunjun Shin; Dexter Hadley
BibTeX | Tags:
@unpublished{nokey,
title = {Knowledge Transfer-Driven Estimation of Knee Moments and Ground Reaction Forces from Smartphone Videos via Temporal-Spatial Modeling of Augmented Joint Dynamics},
author = {Md Sanzid Bin Hossain and Hwan Choi and Zhishan Guo and Sunyong Yoo and Min-Keun Song and Hyunjun Shin and Dexter Hadley},
year = {2025},
date = {2025-12-30},
urldate = {2025-12-30},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
44.
Yunju Song; Myeongjin Kim; Sunyong Yoo
BibTeX | Tags:
@unpublished{nokey,
title = {Tissue-Specific Carcinogenicity Prediction Using Multi-Task Learning on Attention-based Graph Neural Networks},
author = {Yunju Song and Myeongjin Kim and Sunyong Yoo},
year = {2025},
date = {2025-12-30},
urldate = {2025-12-30},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
43.
Subhin Seomun; Myoung Jin Lee; Sunyong Yoo
BibTeX | Tags:
@unpublished{nokey,
title = {MTMM -CYP: Prediction of Cytochrome P450 Isoform Inhibitors through Multi-task Learning with Multiple Molecular Representations},
author = {Subhin Seomun and Myoung Jin Lee and Sunyong Yoo},
year = {2025},
date = {2025-12-30},
urldate = {2025-12-30},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
42.
Hwa Jin Cho; Hyejin Yu; Mingi Kang; Dohyeon Lee; Do Wan Kim; Sung-Min Cho; Glenn Whitman; In Seok Jeong; Sunyong Yoo
BibTeX | Tags:
@unpublished{nokey,
title = {Long-Term Neuropsychiatric Disorders and Healthcare Burden Among Survivors of Extracorporeal Life Support: A Nationwide Population-based Cohort Study},
author = {Hwa Jin Cho and Hyejin Yu and Mingi Kang and Dohyeon Lee and Do Wan Kim and Sung-Min Cho and Glenn Whitman and In Seok Jeong and Sunyong Yoo},
year = {2025},
date = {2025-12-26},
keywords = {},
pubstate = {forthcoming},
tppubtype = {unpublished}
}
41.
DoHyeon Lee; Samel Park; Hyejin Yu; Eunjung Cho; Seung Seok Han; Eun Sil Koh; Byung Ha Chung; Kyung Hwan Jeong; Soo Jeong Choi; Eun Young Lee; Su Hyun Kim; Eun Hui Bae; Sunyong Yoo; Young Joo Kwon
Abstract | Links | BibTeX | Dimensions | Tags: Medical informatics, National health insurance service
@article{Lee2025b,
title = {Current treatment status of fabry disease in South Korea: a longitudinal National health insurance service data-based study},
author = {DoHyeon Lee and Samel Park and Hyejin Yu and Eunjung Cho and Seung Seok Han and Eun Sil Koh and Byung Ha Chung and Kyung Hwan Jeong and Soo Jeong Choi and Eun Young Lee and Su Hyun Kim and Eun Hui Bae and Sunyong Yoo and Young Joo Kwon
},
url = {https://link.springer.com/article/10.1186/s13023-025-03863-5?utm_source=rct_congratemailt&utm_medium=email&utm_campaign=oa_20250710&utm_content=10.1186/s13023-025-03863-5},
doi = {10.1186/s13023-025-03863-5},
issn = {1750-1172},
year = {2025},
date = {2025-07-10},
urldate = {2025-07-10},
journal = {Orphanet Journal of Rare Diseases},
volume = {20},
number = {355},
abstract = {Background
Fabry disease (FD) is an X-linked lysosomal storage disease caused by a mutation of the gene that encodes the α-galactosidase A enzyme. Treatment for FD is based on an enzyme replacement therapy (ERT), such as agalsidase-β, agalsidase-α, and migalastat. However, studies analyzing effects and outcomes of ERT in FD patients in South Korea are limited.
Materials and methods
Treatment status and clinical outcomes of patients with FD in South Korea were investigated using data from the National Health Insurance Service (NHIS). The NHIS provides a comprehensive range of data across the entire Korean population, enabling an in-depth analysis of clinical outcomes associated with FD, including coronary composite heart disease, cerebrovascular disease, end-stage kidney disease (ESKD).
Results
A total of 228 patients with FD were discovered. The diagnosis was earlier in males (n = 120) than in females (n = 108). Almost 90% of patients were treated only with intravenous agalsidase-β or -α. A total of 15 patients switched from agalsidase to migalastat. All clinical outcomes manifested at an earlier age in males than in females. Particularly, ESKD was more prevalent in males, both before and after diagnosis of FD. Patients who had ESKD at the time of FD diagnosis exhibited a higher hazard ratio (HR) for mortality (HR: 5.01, 95% confidence interval: 1.44–17.46).
Conclusions
Our study showed the current treatment status and clinical outcomes in patients with FD in South Korea. Prior to the diagnosis of FD, a considerable number of patients had already reached ESKD, suggesting a lack of awareness of FD among clinicians. Given the higher mortality rate observed in patients with FD and accompanying ESKD, the necessity to improve awareness of FD is highlighted to facilitate early diagnosis.},
note = {Correspondence to Sunyong Yoo and Young Joo Kwon},
keywords = {Medical informatics, National health insurance service},
pubstate = {published},
tppubtype = {article}
}
Background
Fabry disease (FD) is an X-linked lysosomal storage disease caused by a mutation of the gene that encodes the α-galactosidase A enzyme. Treatment for FD is based on an enzyme replacement therapy (ERT), such as agalsidase-β, agalsidase-α, and migalastat. However, studies analyzing effects and outcomes of ERT in FD patients in South Korea are limited.
Materials and methods
Treatment status and clinical outcomes of patients with FD in South Korea were investigated using data from the National Health Insurance Service (NHIS). The NHIS provides a comprehensive range of data across the entire Korean population, enabling an in-depth analysis of clinical outcomes associated with FD, including coronary composite heart disease, cerebrovascular disease, end-stage kidney disease (ESKD).
Results
A total of 228 patients with FD were discovered. The diagnosis was earlier in males (n = 120) than in females (n = 108). Almost 90% of patients were treated only with intravenous agalsidase-β or -α. A total of 15 patients switched from agalsidase to migalastat. All clinical outcomes manifested at an earlier age in males than in females. Particularly, ESKD was more prevalent in males, both before and after diagnosis of FD. Patients who had ESKD at the time of FD diagnosis exhibited a higher hazard ratio (HR) for mortality (HR: 5.01, 95% confidence interval: 1.44–17.46).
Conclusions
Our study showed the current treatment status and clinical outcomes in patients with FD in South Korea. Prior to the diagnosis of FD, a considerable number of patients had already reached ESKD, suggesting a lack of awareness of FD among clinicians. Given the higher mortality rate observed in patients with FD and accompanying ESKD, the necessity to improve awareness of FD is highlighted to facilitate early diagnosis.
Fabry disease (FD) is an X-linked lysosomal storage disease caused by a mutation of the gene that encodes the α-galactosidase A enzyme. Treatment for FD is based on an enzyme replacement therapy (ERT), such as agalsidase-β, agalsidase-α, and migalastat. However, studies analyzing effects and outcomes of ERT in FD patients in South Korea are limited.
Materials and methods
Treatment status and clinical outcomes of patients with FD in South Korea were investigated using data from the National Health Insurance Service (NHIS). The NHIS provides a comprehensive range of data across the entire Korean population, enabling an in-depth analysis of clinical outcomes associated with FD, including coronary composite heart disease, cerebrovascular disease, end-stage kidney disease (ESKD).
Results
A total of 228 patients with FD were discovered. The diagnosis was earlier in males (n = 120) than in females (n = 108). Almost 90% of patients were treated only with intravenous agalsidase-β or -α. A total of 15 patients switched from agalsidase to migalastat. All clinical outcomes manifested at an earlier age in males than in females. Particularly, ESKD was more prevalent in males, both before and after diagnosis of FD. Patients who had ESKD at the time of FD diagnosis exhibited a higher hazard ratio (HR) for mortality (HR: 5.01, 95% confidence interval: 1.44–17.46).
Conclusions
Our study showed the current treatment status and clinical outcomes in patients with FD in South Korea. Prior to the diagnosis of FD, a considerable number of patients had already reached ESKD, suggesting a lack of awareness of FD among clinicians. Given the higher mortality rate observed in patients with FD and accompanying ESKD, the necessity to improve awareness of FD is highlighted to facilitate early diagnosis.
40.
Hyejin Yu; Kwanyong Choi; Ji Yeon Kim; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Artificial Intelligence, Bioinformatics, Ethnopharmacology, Herbal medicine, Network analysis
@article{Yu2025,
title = {Multi-level association rule mining and network pharmacology to identify the polypharmacological effects of herbal materials and compounds in traditional medicine},
author = {Hyejin Yu and Kwanyong Choi and Ji Yeon Kim and Sunyong Yoo},
url = {https://academic.oup.com/bib/article/26/4/bbaf328/8190205?utm_source=advanceaccess&utm_campaign=bib&utm_medium=email},
doi = {10.1093/bib/bbaf328},
issn = {1477-4054},
year = {2025},
date = {2025-07-01},
urldate = {2025-07-01},
journal = {Briefings in Bioinformatics},
volume = {26},
issue = {4},
abstract = {Many cultures worldwide have widely used traditional medicine (TM) to prevent or treat diseases. Herbal materials and their compounds used in TM offer many advantages for drug discovery, including cost-effectiveness, fewer side effects, and improved metabolism. However, the multi-compound and multi-target characteristics of TM prescriptions complicate drug discovery; meanwhile, previous studies have been limited by a lack of high-quality data, complex interpretation, and/or narrow analytical ranges. Thus, this study proposed a framework to identify potential therapeutic combinations of herbal materials and their compounds currently used in TM by integrating association rule mining (ARM) and network pharmacology analysis across multiple TM and biological levels. Subsequently, we collected prescriptions, herbal materials, compounds, genes, phenotypes, and all ensuing interactions to identify effective combinations of herbal materials and compounds using ARM for various symptoms and diseases. This proposed analytical approach was also applied to screen effective herbal material combinations and compounds for five phenotypes: asthma, diabetes, arthritis, stroke, and inflammation. The potential pharmacological effects of the inferred candidates were identified at the molecular level using structural network analysis and a literature review. In addition, compounds from Morus alba, Ephedra sinica, Perilla frutescens, and Pinellia ternata, which were strongly associated with asthma, were validated in vitro. Collectively, our study provides ethnopharmacological and biological evidence for the polypharmacological effects of herbal materials and their compounds, thus enhancing the understanding of the mechanisms involved in TM and suggesting potential candidates for prescriptions, dietary supplements, and drug combinations. The source code and results are available at https://github.com/bmil-jnu/InPETM.},
note = {Correspondence to Sunyong Yoo},
keywords = {Artificial Intelligence, Bioinformatics, Ethnopharmacology, Herbal medicine, Network analysis},
pubstate = {published},
tppubtype = {article}
}
Many cultures worldwide have widely used traditional medicine (TM) to prevent or treat diseases. Herbal materials and their compounds used in TM offer many advantages for drug discovery, including cost-effectiveness, fewer side effects, and improved metabolism. However, the multi-compound and multi-target characteristics of TM prescriptions complicate drug discovery; meanwhile, previous studies have been limited by a lack of high-quality data, complex interpretation, and/or narrow analytical ranges. Thus, this study proposed a framework to identify potential therapeutic combinations of herbal materials and their compounds currently used in TM by integrating association rule mining (ARM) and network pharmacology analysis across multiple TM and biological levels. Subsequently, we collected prescriptions, herbal materials, compounds, genes, phenotypes, and all ensuing interactions to identify effective combinations of herbal materials and compounds using ARM for various symptoms and diseases. This proposed analytical approach was also applied to screen effective herbal material combinations and compounds for five phenotypes: asthma, diabetes, arthritis, stroke, and inflammation. The potential pharmacological effects of the inferred candidates were identified at the molecular level using structural network analysis and a literature review. In addition, compounds from Morus alba, Ephedra sinica, Perilla frutescens, and Pinellia ternata, which were strongly associated with asthma, were validated in vitro. Collectively, our study provides ethnopharmacological and biological evidence for the polypharmacological effects of herbal materials and their compounds, thus enhancing the understanding of the mechanisms involved in TM and suggesting potential candidates for prescriptions, dietary supplements, and drug combinations. The source code and results are available at https://github.com/bmil-jnu/InPETM.
39.
Junyong Park; Hwa-Jin Cho; Sunyong Yoo; Mim-Keun Song
Links | BibTeX | Dimensions | Tags:
@article{Park2025,
title = {Characteristics of Children with Disability through Infant and Children’s Health Screening in South Korea},
author = {Junyong Park and Hwa-Jin Cho and Sunyong Yoo and Mim-Keun Song},
doi = {10.1080/07853890.2025.2525401},
year = {2025},
date = {2025-06-30},
urldate = {2025-06-30},
journal = {Annals of Medicine},
note = {Correspondence to Sunyong Yoo},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
38.
Sunwoo Jung; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: ADR, Artificial Intelligence, Attention mechanism, Bioinformatics, DDI, Deep learning, Text mining
@article{Jung2024,
title = {Interpretable prediction of drug-drug interactions via text embedding in biomedical literature},
author = {Sunwoo Jung and Sunyong Yoo},
url = {https://www.sciencedirect.com/science/article/pii/S0010482524015816},
doi = {10.1016/j.compbiomed.2024.109496},
isbn = {0010-4825},
year = {2025},
date = {2025-02-01},
urldate = {2025-02-01},
journal = {Computers in Biology and Medicine},
volume = {185},
pages = {109496},
abstract = {Polypharmacy is a promising approach for treating diseases, especially those with complex symptoms. However, it can lead to unexpected drug-drug interactions (DDIs), potentially reducing efficacy and triggering adverse drug reactions (ADRs). Predicting the risk of DDIs is crucial for ensuring safe drug use, particularly by identifying the types of DDIs and the mechanisms involved. Therefore, this study used biomedical literature to proposed hierarchical attention-based deep learning models to predict DDIs and their types. The proposed model consists of two components: drug embedding and DDI prediction. The drug embedding module extracts representation vectors that effectively capture drug properties using sentence and sequence embedding methods. For sentence embedding, a pre-trained biomedical language model is used to map drug-related sentences into vector space. For sequence embedding, sentence embedding vectors are sequentially fed into bidirectional long short-term memory with a hierarchical attention network, enabling the analysis of sentences relevant to DDI prediction while accounting for the order of the sentences. Finally, DDI prediction is performed using a deep neural network based on the sequence embedding vectors of a drug pair. Our model achieved high performances in the accuracy (0.85–0.90), AUROC (0.98–0.99), and AUPR (0.63–0.95) performance across 164 DDI types. Additionally, the proposed model showed improvements in up to 11 % in AUROC, and 8 % in AUPR. Furthermore, model interprets predictions by leveraging attention mechanisms and drug similarity. The results indicated that the model considered various factors beyond similarity to predict DDIs. These findings may help prevent unforeseen medical accidents and reduce healthcare costs by predicting detailed drug interaction types.},
note = {Correspondence to Sunyong Yoo},
keywords = {ADR, Artificial Intelligence, Attention mechanism, Bioinformatics, DDI, Deep learning, Text mining},
pubstate = {published},
tppubtype = {article}
}
Polypharmacy is a promising approach for treating diseases, especially those with complex symptoms. However, it can lead to unexpected drug-drug interactions (DDIs), potentially reducing efficacy and triggering adverse drug reactions (ADRs). Predicting the risk of DDIs is crucial for ensuring safe drug use, particularly by identifying the types of DDIs and the mechanisms involved. Therefore, this study used biomedical literature to proposed hierarchical attention-based deep learning models to predict DDIs and their types. The proposed model consists of two components: drug embedding and DDI prediction. The drug embedding module extracts representation vectors that effectively capture drug properties using sentence and sequence embedding methods. For sentence embedding, a pre-trained biomedical language model is used to map drug-related sentences into vector space. For sequence embedding, sentence embedding vectors are sequentially fed into bidirectional long short-term memory with a hierarchical attention network, enabling the analysis of sentences relevant to DDI prediction while accounting for the order of the sentences. Finally, DDI prediction is performed using a deep neural network based on the sequence embedding vectors of a drug pair. Our model achieved high performances in the accuracy (0.85–0.90), AUROC (0.98–0.99), and AUPR (0.63–0.95) performance across 164 DDI types. Additionally, the proposed model showed improvements in up to 11 % in AUROC, and 8 % in AUPR. Furthermore, model interprets predictions by leveraging attention mechanisms and drug similarity. The results indicated that the model considered various factors beyond similarity to predict DDIs. These findings may help prevent unforeseen medical accidents and reduce healthcare costs by predicting detailed drug interaction types.
37.
Dohyeon Lee; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Artificial Intelligence, Attention mechanism, Bioinformatics, Cardiotoxicity, Deep learning, Graph attention network
@article{Lee2025,
title = {hERGAT: predicting hERG blockers using graph attention mechanism through atom- and molecule-level interaction analyses},
author = {Dohyeon Lee and Sunyong Yoo},
url = {https://link.springer.com/article/10.1186/s13321-025-00957-x?utm_source=rct_congratemailt&utm_medium=email&utm_campaign=oa_20250128&utm_content=10.1186/s13321-025-00957-x},
doi = {10.1186/s13321-025-00957-x},
issn = {1758-2946},
year = {2025},
date = {2025-01-28},
urldate = {2025-01-28},
journal = {Journal of Cheminformatics},
volume = {17},
number = {11},
abstract = {The human ether-a-go-go-related gene (hERG) channel plays a critical role in the electrical activity of the heart, and its blockers can cause serious cardiotoxic effects. Thus, screening for hERG channel blockers is a crucial step in the drug development process. Many in silico models have been developed to predict hERG blockers, which can efficiently save time and resources. However, previous methods have found it hard to achieve high performance and to interpret the predictive results. To overcome these challenges, we have proposed hERGAT, a graph neural network model with an attention mechanism, to consider compound interactions on atomic and molecular levels. In the atom-level interaction analysis, we applied a graph attention mechanism (GAT) that integrates information from neighboring nodes and their extended connections. The hERGAT employs a gated recurrent unit (GRU) with the GAT to learn information between more distant atoms. To confirm this, we performed clustering analysis and visualized a correlation heatmap, verifying the interactions between distant atoms were considered during the training process. In the molecule-level interaction analysis, the attention mechanism enables the target node to focus on the most relevant information, highlighting the molecular substructures that play crucial roles in predicting hERG blockers. Through a literature review, we confirmed that highlighted substructures have a significant role in determining the chemical and biological characteristics related to hERG activity. Furthermore, we integrated physicochemical properties into our hERGAT model to improve the performance. Our model achieved an area under the receiver operating characteristic of 0.907 and an area under the precision-recall of 0.904, demonstrating its effectiveness in modeling hERG activity and offering a reliable framework for optimizing drug safety in early development stages.},
note = {Correspondence to Sunyong Yoo},
keywords = {Artificial Intelligence, Attention mechanism, Bioinformatics, Cardiotoxicity, Deep learning, Graph attention network},
pubstate = {published},
tppubtype = {article}
}
The human ether-a-go-go-related gene (hERG) channel plays a critical role in the electrical activity of the heart, and its blockers can cause serious cardiotoxic effects. Thus, screening for hERG channel blockers is a crucial step in the drug development process. Many in silico models have been developed to predict hERG blockers, which can efficiently save time and resources. However, previous methods have found it hard to achieve high performance and to interpret the predictive results. To overcome these challenges, we have proposed hERGAT, a graph neural network model with an attention mechanism, to consider compound interactions on atomic and molecular levels. In the atom-level interaction analysis, we applied a graph attention mechanism (GAT) that integrates information from neighboring nodes and their extended connections. The hERGAT employs a gated recurrent unit (GRU) with the GAT to learn information between more distant atoms. To confirm this, we performed clustering analysis and visualized a correlation heatmap, verifying the interactions between distant atoms were considered during the training process. In the molecule-level interaction analysis, the attention mechanism enables the target node to focus on the most relevant information, highlighting the molecular substructures that play crucial roles in predicting hERG blockers. Through a literature review, we confirmed that highlighted substructures have a significant role in determining the chemical and biological characteristics related to hERG activity. Furthermore, we integrated physicochemical properties into our hERGAT model to improve the performance. Our model achieved an area under the receiver operating characteristic of 0.907 and an area under the precision-recall of 0.904, demonstrating its effectiveness in modeling hERG activity and offering a reliable framework for optimizing drug safety in early development stages.
36.
송윤주; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Graph attention network
@article{nokey,
title = {단일 분자화합물의 폐 발암성 예측을 위한 그래프 신경망 접근법},
author = {송윤주 and 유선용},
url = {https://www.dbpia.co.kr/pdf/pdfView.do?nodeId=NODE12252213},
doi = {10.5626/JOK.2025.52.6.482},
year = {2025},
date = {2025-01-02},
urldate = {2025-01-02},
journal = {정보과학회논문지},
volume = {25},
number = {6},
pages = {482-489},
abstract = {암은 전 세계적으로 매년 수백만 명의 사망자를 초래하는 주요 질환 중 하나로, 특히 폐암은 2022년 한국에서 암 중 가장 높은 사망률을 기록했다. 이에 따라 폐암을 유발하는 화합물에 대한 연구가 필수적이며, 본 연구는 기존 기계학습 및 딥러닝 방법의 한계를 극복하고, 그래프 신경망을 활용하여 폐암유발 가능성을 예측하는 새로운 접근방식을 제안하고 평가했다. 화합물 발암성 데이터베이스인 CPDB, CCRIS, IRIS, T3DB의 SMILES(Simplified Molecular Input Line Entry System) 정보를 기반으로 분자의 구조와 화학적 성질을 그래프 데이터로 변환해 학습했으며, 제안된 모델은 다른 모델 대비 우수한 예측 성능을 보였다. 이는 폐암 예측에 효과적인 도구로서 그래프 신경망의 잠재력을 입증하며, 향후 암 연구와 치료 개발에 중요한 기여를 할 수 있음을 시사한다.},
keywords = {Bioinformatics, Graph attention network},
pubstate = {published},
tppubtype = {article}
}
암은 전 세계적으로 매년 수백만 명의 사망자를 초래하는 주요 질환 중 하나로, 특히 폐암은 2022년 한국에서 암 중 가장 높은 사망률을 기록했다. 이에 따라 폐암을 유발하는 화합물에 대한 연구가 필수적이며, 본 연구는 기존 기계학습 및 딥러닝 방법의 한계를 극복하고, 그래프 신경망을 활용하여 폐암유발 가능성을 예측하는 새로운 접근방식을 제안하고 평가했다. 화합물 발암성 데이터베이스인 CPDB, CCRIS, IRIS, T3DB의 SMILES(Simplified Molecular Input Line Entry System) 정보를 기반으로 분자의 구조와 화학적 성질을 그래프 데이터로 변환해 학습했으며, 제안된 모델은 다른 모델 대비 우수한 예측 성능을 보였다. 이는 폐암 예측에 효과적인 도구로서 그래프 신경망의 잠재력을 입증하며, 향후 암 연구와 치료 개발에 중요한 기여를 할 수 있음을 시사한다.
35.
박준영; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Drugs, Transformer
@article{박준영;유선용2025,
title = {화합물의 골격구조를 활용한 Transformer 기반 새로운 분자 설계},
author = {박준영 and 유선용},
url = {http://journal.dcs.or.kr/_common/do.php?a=full&b=12&bidx=3950&aidx=43776},
doi = {10.9728/dcs.2025.26.1.217},
issn = {1598-2009},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {디지털콘텐츠학회논문지},
volume = {26},
number = {1},
pages = {217-223},
abstract = {전통적인 신약 개발은 새로운 약물을 시장에 출시하기까지 많은 시간과 막대한 비용이 소요되며, 높은 실패율로 인해 효율성이 낮다는 문제가 있다. 이러한 문제를 해결하기 위해 생성 모델을 활용한 혁신적인 접근법이 주목받고 있다. 본 연구에서는 트랜스포머 디코더 구조를 기반으로 화합물의 구조 정보를 문자열로 학습하여 새로운 화합물 구조를 생성하는 모델을 제안한다. 특히, 화합물에서 추출한 골격 구조(scaffold)를 임베딩하여 모델 입력에 포함함으로써, 결합 및 원자 정보와 골격 구조를 동시에 처리하였다. 벤치마크 데이터셋을 사용한 평가 결과, 골격 구조 임베딩을 적용한 모델이 데이터셋 별로 유효성 지표에서 0.964, 0.986의 우수한 성능을 보였다. 본 연구는 분자 생성 모델에 골격 구조 임베딩을 도입함으로써, 화학적 규칙을 준수하는 분자를 효과적으로 생성할 수 있는 방법을 제시하였으며, 신약 개발 분야에서 AI 기반 분자 설계의 효율성을 높이는 데 기여할 것으로 기대된다.},
keywords = {Bioinformatics, Drugs, Transformer},
pubstate = {published},
tppubtype = {article}
}
전통적인 신약 개발은 새로운 약물을 시장에 출시하기까지 많은 시간과 막대한 비용이 소요되며, 높은 실패율로 인해 효율성이 낮다는 문제가 있다. 이러한 문제를 해결하기 위해 생성 모델을 활용한 혁신적인 접근법이 주목받고 있다. 본 연구에서는 트랜스포머 디코더 구조를 기반으로 화합물의 구조 정보를 문자열로 학습하여 새로운 화합물 구조를 생성하는 모델을 제안한다. 특히, 화합물에서 추출한 골격 구조(scaffold)를 임베딩하여 모델 입력에 포함함으로써, 결합 및 원자 정보와 골격 구조를 동시에 처리하였다. 벤치마크 데이터셋을 사용한 평가 결과, 골격 구조 임베딩을 적용한 모델이 데이터셋 별로 유효성 지표에서 0.964, 0.986의 우수한 성능을 보였다. 본 연구는 분자 생성 모델에 골격 구조 임베딩을 도입함으로써, 화학적 규칙을 준수하는 분자를 효과적으로 생성할 수 있는 방법을 제시하였으며, 신약 개발 분야에서 AI 기반 분자 설계의 효율성을 높이는 데 기여할 것으로 기대된다.
2024
34.
Kwanyong Choi; Soyeon Lee; Sunyong Yoo; Hyoung-Yun Han; Soo-yeon Park; Ji Yeon Kim
Abstract | Links | BibTeX | Dimensions | Tags: Drug-induced liver injury, in silico, in vitro
@article{nokeye,
title = {Prediction of bioactive compounds hepatotoxicity using in silico and in vitro analysis},
author = {Kwanyong Choi and Soyeon Lee and Sunyong Yoo and Hyoung-Yun Han and Soo-yeon Park and Ji Yeon Kim},
doi = {10.1186/s13765-024-00961-z},
year = {2024},
date = {2024-12-17},
urldate = {2024-12-17},
journal = {Applied Biological Chemistry},
volume = {67},
number = {107},
abstract = {The leading safety issue and side effect associated with natural herb products is drug-induced liver injury (DILI) caused by bioactive compounds derived from the herb products. Herein, in silico and in vitro analyses were compared to determine the hepatotoxicity of compounds. The results of in silico analyses, which included an integrated database and an interpretable DILI prediction model, identified calycosin, biochanin_A, xanthatin, piperine, and atractyloside as potential hepatotoxic compounds and tenuifolin as a non-hepatotoxic compound. To evaluate the viability of HepG2 cells exposed to the selected compounds, we determined the IC50 and IC20 values of viability using MTT assays. For in-depth screening, we performed hematoxylin and eosin-stained morphological screens, JC-1 mitochondrial assays, and mRNA microarrays. The results indicated that calycosin, biochanin_A, xanthatin, piperine, and atractyloside were potential hepatotoxicants that caused decreased viability and an apoptotic phase in morphology, while these effects were not observed for tenuifolin, a non-hepatotoxicant. In the JC-1 assay, apoptosis was induced by all the predicted hepatotoxicants except atractyloside. According to transcriptomic analysis, all the compounds predicted to induce DILI showed hepatotoxic effects. These results highlighted the importance of using in vitro assays to validate predictive in silico models and determine the potential of bioactive compounds to induce hepatotoxicity in humans.},
note = {Correspondence to Ji Yeon Kim},
keywords = {Drug-induced liver injury, in silico, in vitro},
pubstate = {published},
tppubtype = {article}
}
The leading safety issue and side effect associated with natural herb products is drug-induced liver injury (DILI) caused by bioactive compounds derived from the herb products. Herein, in silico and in vitro analyses were compared to determine the hepatotoxicity of compounds. The results of in silico analyses, which included an integrated database and an interpretable DILI prediction model, identified calycosin, biochanin_A, xanthatin, piperine, and atractyloside as potential hepatotoxic compounds and tenuifolin as a non-hepatotoxic compound. To evaluate the viability of HepG2 cells exposed to the selected compounds, we determined the IC50 and IC20 values of viability using MTT assays. For in-depth screening, we performed hematoxylin and eosin-stained morphological screens, JC-1 mitochondrial assays, and mRNA microarrays. The results indicated that calycosin, biochanin_A, xanthatin, piperine, and atractyloside were potential hepatotoxicants that caused decreased viability and an apoptotic phase in morphology, while these effects were not observed for tenuifolin, a non-hepatotoxicant. In the JC-1 assay, apoptosis was induced by all the predicted hepatotoxicants except atractyloside. According to transcriptomic analysis, all the compounds predicted to induce DILI showed hepatotoxic effects. These results highlighted the importance of using in vitro assays to validate predictive in silico models and determine the potential of bioactive compounds to induce hepatotoxicity in humans.
33.
Hyeon Jae Lee; Kyeong Jin Kim; Soo-yeon Park; Kwanyong Choi; Jaeho Pyee; Sunyong Yoo; Ji Yeon Kim
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Gut permeability, Inflammatory bowel disease, Network analysis
@article{lee2024enhancing,
title = {Enhancing intestinal health with germinated oats: Bioinformatics and compound profiling insights into a novel approach for managing inflammatory bowel disease},
author = {Hyeon Jae Lee and Kyeong Jin Kim and Soo-yeon Park and Kwanyong Choi and Jaeho Pyee and Sunyong Yoo and Ji Yeon Kim},
url = {https://www.sciencedirect.com/science/article/pii/S221242922401263X},
doi = {10.1016/j.fbio.2024.104833},
year = {2024},
date = {2024-10-01},
urldate = {2024-10-01},
journal = {Food Bioscience},
volume = {61},
pages = {104833},
publisher = {Elsevier},
abstract = {Oats are widely recognized for their numerous health benefits, particularly regarding their anti-inflammatory properties. However, research exploring their specific effects on intestinal permeability and tight junction (TJ) integrity in the context of inflammatory bowel disease (IBD) has been limited. This study aimed to investigate the therapeutic efficacy of germinated oat extract (GOE) in managing IBD, a condition marked by persistent gastrointestinal inflammation and increasing global prevalence. The identified compounds were used to predict target biomarkers and mechanisms related to IBD via bioinformatics analysis and validated using in vitro models. In this study, we used network biology and chemical informatics approaches to predict target biomarkers and their molecular mechanisms. The predicted biomarkers were validated for their effectiveness using a cellular model of intestinal inflammation. The effectiveness of treatment with GOE was validated via in vitro studies, which demonstrated significant enhancement in transepithelial electrical resistance (TEER) and a reduction in fluorescein isothiocyanate (FITC) permeability. Analysis of the mRNA expression of IBD-associated biomarkers in Caco-2 cells demonstrated a significant increase in the mRNA levels of TJ proteins, including TJP1, TJP2, occludin, claudin-1 and claudin-3 compared to the inflammatory group. Furthermore, treatment with GOE markedly reduced the mRNA expression levels of proinflammatory cytokines such as TNF-α, IL-6, and CXCL8. The combination of COCONUT and chemical profiling analysis provided insights into the fundamental molecular mechanisms of GOE. These results underscore the potential of systematically using big data-driven network biology to analyze the effect of food components, highlighting GOE as a promising dietary intervention for IBD.},
note = {Correspondence to Ji Yeon Kim},
keywords = {Bioinformatics, Gut permeability, Inflammatory bowel disease, Network analysis},
pubstate = {published},
tppubtype = {article}
}
Oats are widely recognized for their numerous health benefits, particularly regarding their anti-inflammatory properties. However, research exploring their specific effects on intestinal permeability and tight junction (TJ) integrity in the context of inflammatory bowel disease (IBD) has been limited. This study aimed to investigate the therapeutic efficacy of germinated oat extract (GOE) in managing IBD, a condition marked by persistent gastrointestinal inflammation and increasing global prevalence. The identified compounds were used to predict target biomarkers and mechanisms related to IBD via bioinformatics analysis and validated using in vitro models. In this study, we used network biology and chemical informatics approaches to predict target biomarkers and their molecular mechanisms. The predicted biomarkers were validated for their effectiveness using a cellular model of intestinal inflammation. The effectiveness of treatment with GOE was validated via in vitro studies, which demonstrated significant enhancement in transepithelial electrical resistance (TEER) and a reduction in fluorescein isothiocyanate (FITC) permeability. Analysis of the mRNA expression of IBD-associated biomarkers in Caco-2 cells demonstrated a significant increase in the mRNA levels of TJ proteins, including TJP1, TJP2, occludin, claudin-1 and claudin-3 compared to the inflammatory group. Furthermore, treatment with GOE markedly reduced the mRNA expression levels of proinflammatory cytokines such as TNF-α, IL-6, and CXCL8. The combination of COCONUT and chemical profiling analysis provided insights into the fundamental molecular mechanisms of GOE. These results underscore the potential of systematically using big data-driven network biology to analyze the effect of food components, highlighting GOE as a promising dietary intervention for IBD.
32.
Suyeon Kim; Dong Young Kim; Je Won Park; Shinwook Kim; Seungchan Lee; Han Seung Jang; Jinseok Park; Sunyong Yoo; Myoung Jin Lee
Abstract | Links | BibTeX | Dimensions | Tags: Optimization
@article{kim2024passing,
title = {Passing Word Line-Induced Subthreshold Leakage Reduction Using a Partial Insulator in a Buried Channel Array Transistor},
author = {Suyeon Kim and Dong Young Kim and Je Won Park and Shinwook Kim and Seungchan Lee and Han Seung Jang and Jinseok Park and Sunyong Yoo and Myoung Jin Lee},
url = {https://ieeexplore.ieee.org/abstract/document/10495758},
doi = {10.1109/TED.2024.3379963},
issn = {0018-9383},
year = {2024},
date = {2024-04-10},
urldate = {2024-04-10},
journal = {IEEE Transactions on Electron Devices},
volume = {71},
issue = {5},
pages = {2976 - 2982},
publisher = {IEEE},
abstract = {As dynamic random access memory (DRAM) technologies continue to be downscaled, the partial isolation type buried channel array transistor (Pi-BCAT) structure has emerged as an innovative solution for the increasing challenges caused by leakage current adjacent to passing word lines (PWLs). This study reveals that the Pi-BCAT reduces leakage currents by 30% when compared to conventional BCAT structures. Our comprehensive simulations demonstrate that Pi-BCAT is resistant to temperature-induced leakage variations, confirming its significance in promoting consistent device performance and power management. The Pi-BCAT structure is predicted to be crucial in the advancement of DRAM reliability and efficiency, hence initiating further advancements in semiconductor technology.},
note = {Correspondence to Sunyong Yoo and Myoung Jin Lee},
keywords = {Optimization},
pubstate = {published},
tppubtype = {article}
}
As dynamic random access memory (DRAM) technologies continue to be downscaled, the partial isolation type buried channel array transistor (Pi-BCAT) structure has emerged as an innovative solution for the increasing challenges caused by leakage current adjacent to passing word lines (PWLs). This study reveals that the Pi-BCAT reduces leakage currents by 30% when compared to conventional BCAT structures. Our comprehensive simulations demonstrate that Pi-BCAT is resistant to temperature-induced leakage variations, confirming its significance in promoting consistent device performance and power management. The Pi-BCAT structure is predicted to be crucial in the advancement of DRAM reliability and efficiency, hence initiating further advancements in semiconductor technology.
31.
Myeonghyeon Jeong; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Attention mechanism, Bioinformatics, Deep learning, Fetotoxicity, in silico, Interpretability
@article{jeong2024fetoml,
title = {FetoML: Interpretable predictions of the fetotoxicity of drugs based on machine learning approaches},
author = {Myeonghyeon Jeong and Sunyong Yoo},
url = {https://onlinelibrary.wiley.com/doi/full/10.1002/minf.202300312},
doi = {10.1002/minf.202300312},
issn = {1868-1743},
year = {2024},
date = {2024-03-03},
urldate = {2024-03-03},
journal = {Molecular Informatics},
volume = {43},
number = {6},
pages = {e202300312},
publisher = {Wiley Online Library},
abstract = {Pregnant females may use medications to manage health problems that develop during pregnancy or that they had prior to pregnancy. However, using medications during pregnancy has a potential risk to the fetus. Assessing the fetotoxicity of drugs is essential to ensure safe treatments, but the current process is challenged by ethical issues, time, and cost. Therefore, the need for in silico models to efficiently assess the fetotoxicity of drugs has recently emerged. Previous studies have proposed successful machine learning models for fetotoxicity prediction and even suggest molecular substructures that are possibly associated with fetotoxicity risks or protective effects. However, the interpretation of the decisions of the models on fetotoxicity prediction for each drug is still insufficient. This study constructed machine learning-based models that can predict the fetotoxicity of drugs while providing explanations for the decisions. For this, permutation feature importance was used to identify the general features that the model made significant in predicting the fetotoxicity of drugs. In addition, features associated with fetotoxicity for each drug were analyzed using the attention mechanism. The predictive performance of all the constructed models was significantly high (AUROC: 0.854-0.974, AUPR: 0.890-0.975). Furthermore, we conducted literature reviews on the predicted important features and found that they were highly associated with fetotoxicity. We expect that our model will benefit fetotoxicity research by providing an evaluation of fetotoxicity risks for drugs or drug candidates, along with an interpretation of that prediction.},
note = {Correspondence to Sunyong Yoo},
keywords = {Attention mechanism, Bioinformatics, Deep learning, Fetotoxicity, in silico, Interpretability},
pubstate = {published},
tppubtype = {article}
}
Pregnant females may use medications to manage health problems that develop during pregnancy or that they had prior to pregnancy. However, using medications during pregnancy has a potential risk to the fetus. Assessing the fetotoxicity of drugs is essential to ensure safe treatments, but the current process is challenged by ethical issues, time, and cost. Therefore, the need for in silico models to efficiently assess the fetotoxicity of drugs has recently emerged. Previous studies have proposed successful machine learning models for fetotoxicity prediction and even suggest molecular substructures that are possibly associated with fetotoxicity risks or protective effects. However, the interpretation of the decisions of the models on fetotoxicity prediction for each drug is still insufficient. This study constructed machine learning-based models that can predict the fetotoxicity of drugs while providing explanations for the decisions. For this, permutation feature importance was used to identify the general features that the model made significant in predicting the fetotoxicity of drugs. In addition, features associated with fetotoxicity for each drug were analyzed using the attention mechanism. The predictive performance of all the constructed models was significantly high (AUROC: 0.854-0.974, AUPR: 0.890-0.975). Furthermore, we conducted literature reviews on the predicted important features and found that they were highly associated with fetotoxicity. We expect that our model will benefit fetotoxicity research by providing an evaluation of fetotoxicity risks for drugs or drug candidates, along with an interpretation of that prediction.
30.
Sunyong Yoo; Myeonghyeon Jeong; Subhin Seomun; Kiseong Kim; Youngmahn Han
Abstract | Links | BibTeX | Dimensions | Tags: Amino acids, Attention mechanism, Bioinformatics, Coronaviruses, Deep learning, Immune system, Lymphocytes, Predictive models, Proteins, Transformer
@article{yoo2024interpretable,
title = {Interpretable Prediction of SARS-CoV-2 Epitope-specific TCR Recognition Using a Pre-Trained Protein Language Model},
author = {Sunyong Yoo and Myeonghyeon Jeong and Subhin Seomun and Kiseong Kim and Youngmahn Han},
url = {https://ieeexplore.ieee.org/abstract/document/10443062},
doi = {10.1109/TCBB.2024.3368046},
year = {2024},
date = {2024-02-21},
urldate = {2024-02-21},
journal = {IEEE/ACM Transactions on Computational Biology and Bioinformatics},
volume = {21},
issue = {3},
pages = {428-438},
publisher = {IEEE},
abstract = {The emergence of the novel coronavirus, designated as severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has posed a significant threat to public health worldwide. There has been progress in reducing hospitalizations and deaths due to SARS-CoV-2. However, challenges stem from the emergence of SARS-CoV-2 variants, which exhibit high transmission rates, increased disease severity, and the ability to evade humoral immunity. Epitope-specific T-cell receptor (TCR) recognition is key in determining the T-cell immunogenicity for SARS-CoV-2 epitopes. Although several data-driven methods for predicting epitope-specific TCR recognition have been proposed, they remain challenging due to the enormous diversity of TCRs and the lack of available training data. Self-supervised transfer learning has recently been proven useful for extracting information from unlabeled protein sequences, increasing the predictive performance of fine-tuned models, and using a relatively small amount of training data. This study presents a deep-learning model generated by fine-tuning pre-trained protein embeddings from a large corpus of protein sequences. The fine-tuned model showed markedly high predictive performance and outperformed the recent Gaussian process-based prediction model. The output attentions captured by the deep-learning model suggested critical amino acid positions in the SARS-CoV-2 epitope-specific TCRβ sequences that are highly associated with the viral escape of T-cell immune response.},
note = {Correspondence to Sunyong Yoo},
keywords = {Amino acids, Attention mechanism, Bioinformatics, Coronaviruses, Deep learning, Immune system, Lymphocytes, Predictive models, Proteins, Transformer},
pubstate = {published},
tppubtype = {article}
}
The emergence of the novel coronavirus, designated as severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), has posed a significant threat to public health worldwide. There has been progress in reducing hospitalizations and deaths due to SARS-CoV-2. However, challenges stem from the emergence of SARS-CoV-2 variants, which exhibit high transmission rates, increased disease severity, and the ability to evade humoral immunity. Epitope-specific T-cell receptor (TCR) recognition is key in determining the T-cell immunogenicity for SARS-CoV-2 epitopes. Although several data-driven methods for predicting epitope-specific TCR recognition have been proposed, they remain challenging due to the enormous diversity of TCRs and the lack of available training data. Self-supervised transfer learning has recently been proven useful for extracting information from unlabeled protein sequences, increasing the predictive performance of fine-tuned models, and using a relatively small amount of training data. This study presents a deep-learning model generated by fine-tuning pre-trained protein embeddings from a large corpus of protein sequences. The fine-tuned model showed markedly high predictive performance and outperformed the recent Gaussian process-based prediction model. The output attentions captured by the deep-learning model suggested critical amino acid positions in the SARS-CoV-2 epitope-specific TCRβ sequences that are highly associated with the viral escape of T-cell immune response.
29.
Soyeon Lee; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Artificial Intelligence, Attention mechanism, Bioinformatics, Deep learning, Drug-induced liver injury, Feature importance, Hepatotoxicity, in silico
@article{lee2024interdili,
title = {InterDILI: interpretable prediction of drug-induced liver injury through permutation feature importance and attention mechanism},
author = {Soyeon Lee and Sunyong Yoo},
url = {https://link.springer.com/article/10.1186/s13321-023-00796-8},
doi = {10.1186/s13321-023-00796-8},
year = {2024},
date = {2024-01-03},
urldate = {2024-01-03},
journal = {Journal of Cheminformatics},
volume = {16},
number = {1},
pages = {1},
publisher = {Springer},
abstract = {Safety is one of the important factors constraining the distribution of clinical drugs on the market. Drug-induced liver injury (DILI) is the leading cause of safety problems produced by drug side effects. Therefore, the DILI risk of approved drugs and potential drug candidates should be assessed. Currently, in vivo and in vitro methods are used to test DILI risk, but both methods are labor-intensive, time-consuming, and expensive. To overcome these problems, many in silico methods for DILI prediction have been suggested. Previous studies have shown that DILI prediction models can be utilized as prescreening tools, and they achieved a good performance. However, there are still limitations in interpreting the prediction results. Therefore, this study focused on interpreting the model prediction to analyze which features could potentially cause DILI. For this, five publicly available datasets were collected to train and test the model. Then, various machine learning methods were applied using substructure and physicochemical descriptors as inputs and the DILI label as the output. The interpretation of feature importance was analyzed by recognizing the following general-to-specific patterns: (i) identifying general important features of the overall DILI predictions, and (ii) highlighting specific molecular substructures which were highly related to the DILI prediction for each compound. The results indicated that the model not only captured the previously known properties to be related to DILI but also proposed a new DILI potential substructural of physicochemical properties. The models for the DILI prediction achieved an area under the receiver operating characteristic (AUROC) of 0.88–0.97 and an area under the Precision-Recall curve (AUPRC) of 0.81–0.95. From this, we hope the proposed models can help identify the potential DILI risk of drug candidates at an early stage and offer valuable insights for drug development.},
note = {Correspondence to Sunyong Yoo},
keywords = {Artificial Intelligence, Attention mechanism, Bioinformatics, Deep learning, Drug-induced liver injury, Feature importance, Hepatotoxicity, in silico},
pubstate = {published},
tppubtype = {article}
}
Safety is one of the important factors constraining the distribution of clinical drugs on the market. Drug-induced liver injury (DILI) is the leading cause of safety problems produced by drug side effects. Therefore, the DILI risk of approved drugs and potential drug candidates should be assessed. Currently, in vivo and in vitro methods are used to test DILI risk, but both methods are labor-intensive, time-consuming, and expensive. To overcome these problems, many in silico methods for DILI prediction have been suggested. Previous studies have shown that DILI prediction models can be utilized as prescreening tools, and they achieved a good performance. However, there are still limitations in interpreting the prediction results. Therefore, this study focused on interpreting the model prediction to analyze which features could potentially cause DILI. For this, five publicly available datasets were collected to train and test the model. Then, various machine learning methods were applied using substructure and physicochemical descriptors as inputs and the DILI label as the output. The interpretation of feature importance was analyzed by recognizing the following general-to-specific patterns: (i) identifying general important features of the overall DILI predictions, and (ii) highlighting specific molecular substructures which were highly related to the DILI prediction for each compound. The results indicated that the model not only captured the previously known properties to be related to DILI but also proposed a new DILI potential substructural of physicochemical properties. The models for the DILI prediction achieved an area under the receiver operating characteristic (AUROC) of 0.88–0.97 and an area under the Precision-Recall curve (AUPRC) of 0.81–0.95. From this, we hope the proposed models can help identify the potential DILI risk of drug candidates at an early stage and offer valuable insights for drug development.
28.
정선우; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: ADR, DDI, Deep learning, Text mining
@article{정선우2024drug,
title = {Drug-Drug Interaction Prediction Model Based on Deep Learning Using Drug Information Document Embedding},
author = {정선우 and 유선용},
url = {https://www.dbpia.co.kr/pdf/pdfView.do?nodeId=NODE11852157&googleIPSandBox=false&mark=0&minRead=10&ipRange=false&b2cLoginYN=false&icstClss=010000&isPDFSizeAllowed=true&nodeHistoryTotalCnt=2&accessgl=Y&language=ko_KR&hasTopBanner=true},
doi = {10.5626/JOK.2024.51.6.503},
issn = {2833-6296},
year = {2024},
date = {2024-01-02},
urldate = {2024-01-02},
journal = {Journal of KIISE},
volume = {51},
number = {6},
pages = {503–512},
abstract = {다약제는 암, 고혈압, 천식 등 다양한 질병에 대하여 유망한 접근법이다. 일반적으로 병원에 방문하는 환자는 2종 이상의 약물을 처방받는다. 그러나 다약제의 사용은 개별 약물이 목표하는 작용 외에 예상치 못한 상호작용을 유발할 수 있다. 약물 간 상호작용을 사전에 예측하는 것은 안전한 약물 사용을 위한 매우 중요한 과제이다. 본 연구에서는 다약제 사용 시 발생 가능한 약물 간 상호작용 예측을 위해 개별 약물 정보를 포함한 문서를 이용하여 약물을 표현하는 문서 임베딩 기반의 딥러닝 예측 모델을 제안한다. 약물 정보 문서는 DrugBank 데이터를 이용해 약물의 설명, 적응증, 약력학 정보, 작용 기전, 독성 속성을 결합해 구축한다. 그 후 Doc2Vec, BioSentVec 언어 모델을 통해 약물 문서로부터 약물 표현 벡터를 생성한다. 두 약물 표현 벡터는 한 쌍으로 묶여 딥러닝 기반 예측 모델에 입력되고, 해당 모델은 두 약물 간 상호작용을 예측한다. 본 논문에서는 언어 임베딩 모델의 성능 비교, 데이터의 불균형도 조절 등 다양한 조건의 변화에 따른 실험 결과의 차이를 분석하여 약물 간 상호작용 예측을 위한 최적의 모델을 구축하는 것을 목표로 한다. 제안된 모델은 약물 처방 과정, 신약 개발의 임상 과정 등에서 약물간 상호작용 사전 예측을 위하여 활용될 수 있을 것으로 기대된다.},
note = {Correspondence to Sunyong Yoo},
keywords = {ADR, DDI, Deep learning, Text mining},
pubstate = {published},
tppubtype = {article}
}
다약제는 암, 고혈압, 천식 등 다양한 질병에 대하여 유망한 접근법이다. 일반적으로 병원에 방문하는 환자는 2종 이상의 약물을 처방받는다. 그러나 다약제의 사용은 개별 약물이 목표하는 작용 외에 예상치 못한 상호작용을 유발할 수 있다. 약물 간 상호작용을 사전에 예측하는 것은 안전한 약물 사용을 위한 매우 중요한 과제이다. 본 연구에서는 다약제 사용 시 발생 가능한 약물 간 상호작용 예측을 위해 개별 약물 정보를 포함한 문서를 이용하여 약물을 표현하는 문서 임베딩 기반의 딥러닝 예측 모델을 제안한다. 약물 정보 문서는 DrugBank 데이터를 이용해 약물의 설명, 적응증, 약력학 정보, 작용 기전, 독성 속성을 결합해 구축한다. 그 후 Doc2Vec, BioSentVec 언어 모델을 통해 약물 문서로부터 약물 표현 벡터를 생성한다. 두 약물 표현 벡터는 한 쌍으로 묶여 딥러닝 기반 예측 모델에 입력되고, 해당 모델은 두 약물 간 상호작용을 예측한다. 본 논문에서는 언어 임베딩 모델의 성능 비교, 데이터의 불균형도 조절 등 다양한 조건의 변화에 따른 실험 결과의 차이를 분석하여 약물 간 상호작용 예측을 위한 최적의 모델을 구축하는 것을 목표로 한다. 제안된 모델은 약물 처방 과정, 신약 개발의 임상 과정 등에서 약물간 상호작용 사전 예측을 위하여 활용될 수 있을 것으로 기대된다.
27.
이도현; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Cardiotoxicity, Graph attention network
@article{nokey,
title = {메시지 패싱 그래프 기반 딥러닝 모델을 활용한 화합물의 심장독성 예측},
author = {이도현 and 유선용},
url = {https://www.dbpia.co.kr/journal/articleDetail?nodeId=NODE11956044},
doi = {10.9728/dcs.2024.25.10.2961},
isbn = {1598-2009},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
journal = {한국디지털콘텐츠학회},
volume = {25},
number = {10},
pages = {2961-2968},
abstract = {hERG 채널은 심장의 전기 활동에 필수적이며, 이 채널을 차단하는 물질은 심각한 심장 독성 효과를 일으킬 수 있다. 인실리코 예측 모델은 hERG 차단제를 효율적으로 선별할 수 있어 시간과 자원을 절약할 수 있다. 이전 접근법은 예측 결과를 해석하고 분자 구조-기능 관계를 이해하는 데 어렵다. 본 연구에서는 공개 데이터베이스로부터 화합물을 수집하여 12,920개의 데이터셋을 구축 하였다. 화합물의 그래프 구조를 고려하는 그래프 신경망(GNN) 가운데 메시지 패싱 신경망(MPNN)을 활용하여 특징 벡터를 추출하고, 이를 구조적ㆍ물리화학적 특성과 결합하여 최종 hERG 차단제를 예측하였다. 해당 모델은 AUROC는 0.864 (±0.009), AUPR은 0.907 (±0.010)의 성능을 달성하였다. 실험 결과, 제안된 모델은 그래프 특징 벡터를 통합하여 분자 특성을 효과적으로 반영하고 분자 간의 관계를 예측하여 hERG 차단제를 예측할 수 있음을 시사한다. 본 연구는 약물 개발과정에서 예비 도구로 활용되어 심장독성을 조기에 평가할 수 있을 것이다.},
note = {Correspondence to Sunyong Yoo},
keywords = {Bioinformatics, Cardiotoxicity, Graph attention network},
pubstate = {published},
tppubtype = {article}
}
hERG 채널은 심장의 전기 활동에 필수적이며, 이 채널을 차단하는 물질은 심각한 심장 독성 효과를 일으킬 수 있다. 인실리코 예측 모델은 hERG 차단제를 효율적으로 선별할 수 있어 시간과 자원을 절약할 수 있다. 이전 접근법은 예측 결과를 해석하고 분자 구조-기능 관계를 이해하는 데 어렵다. 본 연구에서는 공개 데이터베이스로부터 화합물을 수집하여 12,920개의 데이터셋을 구축 하였다. 화합물의 그래프 구조를 고려하는 그래프 신경망(GNN) 가운데 메시지 패싱 신경망(MPNN)을 활용하여 특징 벡터를 추출하고, 이를 구조적ㆍ물리화학적 특성과 결합하여 최종 hERG 차단제를 예측하였다. 해당 모델은 AUROC는 0.864 (±0.009), AUPR은 0.907 (±0.010)의 성능을 달성하였다. 실험 결과, 제안된 모델은 그래프 특징 벡터를 통합하여 분자 특성을 효과적으로 반영하고 분자 간의 관계를 예측하여 hERG 차단제를 예측할 수 있음을 시사한다. 본 연구는 약물 개발과정에서 예비 도구로 활용되어 심장독성을 조기에 평가할 수 있을 것이다.
2023
26.
Sunyong Yoo; Ja Young Choi; Shin-seung Yang; Seong-Eun Koh; Myeong-Hyeon Jeong; Min-Keun Song
Abstract | Links | BibTeX | Dimensions | Tags: Medical informatics, National health insurance service
@article{yoo2023medical,
title = {Medical service utilization by children with physical or brain disabilities in South Korea},
author = {Sunyong Yoo and Ja Young Choi and Shin-seung Yang and Seong-Eun Koh and Myeong-Hyeon Jeong and Min-Keun Song},
url = {https://link.springer.com/article/10.1186/s12887-023-04309-2},
doi = {10.1186/s12887-023-04309-2},
year = {2023},
date = {2023-09-26},
urldate = {2023-09-26},
journal = {BMC pediatrics},
volume = {23},
number = {1},
pages = {487},
publisher = {Springer},
abstract = {Background
Children with physical or brain disabilities experience several functional impairments and declining health complications that must be considered for adequate medical support. This study investigated the current medical service utilization of children expressing physical or brain disabilities in South Korea by analyzing medical visits, expenses, and comorbidities.
Methods
We used a database linked to the National Rehabilitation Center of South Korea to extract information on medical services utilized by children with physical or brain disabilities, the number of children with a disability, medical visits for each child, medical expenses per visit, total medical treatment cost, copayments by age group, condition severity, and disability type.
Results
Brain disorder comorbidities significantly differed between those with mild and severe disabilities. Visits per child, total medical treatment cost, and copayments were higher in children with severe physical disabilities; however, medical expenses per visit were lower than those with mild disabilities. These parameters were higher in children with severe brain disabilities than in mild cases. Total medical expenses incurred by newborns to three-year-old children with physical disorders were highest due to increased visits per child. However, medical expenses per visit were highest for children aged 13–18.
Conclusion
Medical service utilization varied by age, condition severity, and disability type. Severe cases and older children with potentially fatal comorbidities required additional economic support. Therefore, a healthcare delivery system for children with disabilities should be established to set affordable medical costs and provide comprehensive medical services based on disability type and severity.},
note = {Correspondence to Min-Keun Song},
keywords = {Medical informatics, National health insurance service},
pubstate = {published},
tppubtype = {article}
}
Background
Children with physical or brain disabilities experience several functional impairments and declining health complications that must be considered for adequate medical support. This study investigated the current medical service utilization of children expressing physical or brain disabilities in South Korea by analyzing medical visits, expenses, and comorbidities.
Methods
We used a database linked to the National Rehabilitation Center of South Korea to extract information on medical services utilized by children with physical or brain disabilities, the number of children with a disability, medical visits for each child, medical expenses per visit, total medical treatment cost, copayments by age group, condition severity, and disability type.
Results
Brain disorder comorbidities significantly differed between those with mild and severe disabilities. Visits per child, total medical treatment cost, and copayments were higher in children with severe physical disabilities; however, medical expenses per visit were lower than those with mild disabilities. These parameters were higher in children with severe brain disabilities than in mild cases. Total medical expenses incurred by newborns to three-year-old children with physical disorders were highest due to increased visits per child. However, medical expenses per visit were highest for children aged 13–18.
Conclusion
Medical service utilization varied by age, condition severity, and disability type. Severe cases and older children with potentially fatal comorbidities required additional economic support. Therefore, a healthcare delivery system for children with disabilities should be established to set affordable medical costs and provide comprehensive medical services based on disability type and severity.
Children with physical or brain disabilities experience several functional impairments and declining health complications that must be considered for adequate medical support. This study investigated the current medical service utilization of children expressing physical or brain disabilities in South Korea by analyzing medical visits, expenses, and comorbidities.
Methods
We used a database linked to the National Rehabilitation Center of South Korea to extract information on medical services utilized by children with physical or brain disabilities, the number of children with a disability, medical visits for each child, medical expenses per visit, total medical treatment cost, copayments by age group, condition severity, and disability type.
Results
Brain disorder comorbidities significantly differed between those with mild and severe disabilities. Visits per child, total medical treatment cost, and copayments were higher in children with severe physical disabilities; however, medical expenses per visit were lower than those with mild disabilities. These parameters were higher in children with severe brain disabilities than in mild cases. Total medical expenses incurred by newborns to three-year-old children with physical disorders were highest due to increased visits per child. However, medical expenses per visit were highest for children aged 13–18.
Conclusion
Medical service utilization varied by age, condition severity, and disability type. Severe cases and older children with potentially fatal comorbidities required additional economic support. Therefore, a healthcare delivery system for children with disabilities should be established to set affordable medical costs and provide comprehensive medical services based on disability type and severity.
25.
Jinmyung Jung; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Breast cancer, Feature importance, Gene expression, Machine learning, Metastasis marker
@article{jung2023identification,
title = {Identification of Breast Cancer Metastasis Markers from Gene Expression Profiles Using Machine Learning Approaches},
author = {Jinmyung Jung and Sunyong Yoo},
url = {https://www.mdpi.com/2073-4425/14/9/1820},
doi = {10.3390/genes14091820},
year = {2023},
date = {2023-09-20},
urldate = {2023-09-20},
journal = {Genes},
volume = {14},
number = {9},
pages = {1820},
publisher = {MDPI},
abstract = {Cancer metastasis accounts for approximately 90% of cancer deaths, and elucidating markers in metastasis is the first step in its prevention. To characterize metastasis marker genes (MGs) of breast cancer, XGBoost models that classify metastasis status were trained with gene expression profiles from TCGA. Then, a metastasis score (MS) was assigned to each gene by calculating the inner product between the feature importance and the AUC performance of the models. As a result, 54, 202, and 357 genes with the highest MS were characterized as MGs by empirical p-value cutoffs of 0.001, 0.005, and 0.01, respectively. The three sets of MGs were compared with those from existing metastasis marker databases, which provided significant results in most comparisons (p-value < 0.05). They were also significantly enriched in biological processes associated with breast cancer metastasis. The three MGs, SPPL2C, KRT23, and RGS7, showed highly significant results (p-value < 0.01) in the survival analysis. The MGs that could not be identified by statistical analysis (e.g., GOLM1, ELAVL1, UBP1, and AZGP1), as well as the MGs with the highest MS (e.g., ZNF676, FAM163B, LDOC2, IRF1, and STK40), were verified via the literature. Additionally, we checked how close the MGs were to each other in the protein–protein interaction networks. We expect that the characterized markers will help understand and prevent breast cancer metastasis.},
note = {Correspondence to Sunyong Yoo},
keywords = {Bioinformatics, Breast cancer, Feature importance, Gene expression, Machine learning, Metastasis marker},
pubstate = {published},
tppubtype = {article}
}
Cancer metastasis accounts for approximately 90% of cancer deaths, and elucidating markers in metastasis is the first step in its prevention. To characterize metastasis marker genes (MGs) of breast cancer, XGBoost models that classify metastasis status were trained with gene expression profiles from TCGA. Then, a metastasis score (MS) was assigned to each gene by calculating the inner product between the feature importance and the AUC performance of the models. As a result, 54, 202, and 357 genes with the highest MS were characterized as MGs by empirical p-value cutoffs of 0.001, 0.005, and 0.01, respectively. The three sets of MGs were compared with those from existing metastasis marker databases, which provided significant results in most comparisons (p-value < 0.05). They were also significantly enriched in biological processes associated with breast cancer metastasis. The three MGs, SPPL2C, KRT23, and RGS7, showed highly significant results (p-value < 0.01) in the survival analysis. The MGs that could not be identified by statistical analysis (e.g., GOLM1, ELAVL1, UBP1, and AZGP1), as well as the MGs with the highest MS (e.g., ZNF676, FAM163B, LDOC2, IRF1, and STK40), were verified via the literature. Additionally, we checked how close the MGs were to each other in the protein–protein interaction networks. We expect that the characterized markers will help understand and prevent breast cancer metastasis.
24.
Myeonghyeon Jeong; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Machine learning
@article{jeong2023predicting,
title = {Predicting the Fetotoxicity of Drugs Using Machine Learning},
author = {Myeonghyeon Jeong and Sunyong Yoo},
url = {https://koreascience.kr/article/JAKO202320150261638.page},
doi = {10.5352/JLS.2023.33.6.490},
year = {2023},
date = {2023-01-01},
urldate = {2023-01-01},
journal = {Journal of Life Science},
volume = {33},
number = {6},
pages = {490–497},
publisher = {Korean Society of Life Science},
abstract = {Pregnant women may need to take medications to treat preexisting diseases or diseases that develop during pregnancy. However, some drugs may be fetotoxic and lead to, for example, teratogenicity and growth retardation. Predicting the fetotoxicity of drugs is thus important for the health of the mother and fetus. The fetotoxicity of many drugs has not been established because various challenges hinder the ability of researchers to determine their fetotoxicity. The need exists for in silico-based fetotoxicity assessment models, as they can modernize the testing paradigm, improve predictability, and reduce the use of animals and the costs of fetotoxicity testing. In this study, we collected data on the fetotoxicity of drugs and constructed fetotoxicity prediction models based on various machine learning algorithms. We optimized the models for more precise predictions by tuning the hyperparameters. We then performed quantitative performance evaluations. The results indicated that the constructed machine learning-based models had high performance (AUROC >0.85, AUPR >0.9) in fetotoxicity prediction. We also analyzed the feature importance of our model's predictions, which could be leveraged to identify the specific features of drugs that are strongly associated with fetotoxicity. The proposed model can be used to prescreen drugs and drug candidates at a lower cost and in less time. It provides a predictive score for fetotoxicity risk, which may be beneficial in the design of studies on fetotoxicity in human pregnancy.},
note = {Correspondence to Sunyong Yoo},
keywords = {Machine learning},
pubstate = {published},
tppubtype = {article}
}
Pregnant women may need to take medications to treat preexisting diseases or diseases that develop during pregnancy. However, some drugs may be fetotoxic and lead to, for example, teratogenicity and growth retardation. Predicting the fetotoxicity of drugs is thus important for the health of the mother and fetus. The fetotoxicity of many drugs has not been established because various challenges hinder the ability of researchers to determine their fetotoxicity. The need exists for in silico-based fetotoxicity assessment models, as they can modernize the testing paradigm, improve predictability, and reduce the use of animals and the costs of fetotoxicity testing. In this study, we collected data on the fetotoxicity of drugs and constructed fetotoxicity prediction models based on various machine learning algorithms. We optimized the models for more precise predictions by tuning the hyperparameters. We then performed quantitative performance evaluations. The results indicated that the constructed machine learning-based models had high performance (AUROC >0.85, AUPR >0.9) in fetotoxicity prediction. We also analyzed the feature importance of our model's predictions, which could be leveraged to identify the specific features of drugs that are strongly associated with fetotoxicity. The proposed model can be used to prescreen drugs and drug candidates at a lower cost and in less time. It provides a predictive score for fetotoxicity risk, which may be beneficial in the design of studies on fetotoxicity in human pregnancy.
23.
이소연; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Hepatotoxicity, Machine learning
@article{이소연2023기계학습을,
title = {기계학습을 활용한 화합물의 약인성 간 손상 예측 방법 연구},
author = {이소연 and 유선용},
url = {https://www.dbpia.co.kr/pdf/pdfView.do?nodeId=NODE11519759&googleIPSandBox=false&mark=0&minRead=10&ipRange=false&b2cLoginYN=false&icstClss=010000&isPDFSizeAllowed=true&nodeHistoryTotalCnt=2&accessgl=Y&language=ko_KR&hasTopBanner=true},
doi = {10.5626/JOK.2023.50.9.777},
issn = {2383-6296},
year = {2023},
date = {2023-01-01},
urldate = {2023-01-01},
journal = {정보과학회논문지},
volume = {50},
number = {9},
pages = {777–783},
abstract = {약 약인성 간 손상은 임상시험용 의약품이 시장에 유통되는 것을 막는 요인 중 하나이다. 따라서 사전에 화합물의 약인성 간 손상 위험 평가가 필요하다. 안전성을 평가하기 위해 생체 내 (in vivo) 및 시험관 내 시험 방법(in vitro)이 사용되지만 이들은 시간과 비용이 많이 든다. 본 연구에서는 위의 문제를 극복하고자 random forest, light gradient boosting machine, logistic regression 모델을 제안한다. 모델은 입력으로 화합물의 분자 구조와 물리화학적 특징을 사용하고 출력으로 약인성 간 손상을 예측한다. 최적의 모델은 평가 지표에서 전반적으로 좋은 성능을 보인 random forest였다. 본 연구에서 제안된 모델은 신약 후보물질의 잠재적인 간 손상을 미리 파악함으로써 신약 개발 과정에 도움을 줄 수 있을 것으로 기대된다.},
note = {Correspondence to Sunyong Yoo},
keywords = {Hepatotoxicity, Machine learning},
pubstate = {published},
tppubtype = {article}
}
약 약인성 간 손상은 임상시험용 의약품이 시장에 유통되는 것을 막는 요인 중 하나이다. 따라서 사전에 화합물의 약인성 간 손상 위험 평가가 필요하다. 안전성을 평가하기 위해 생체 내 (in vivo) 및 시험관 내 시험 방법(in vitro)이 사용되지만 이들은 시간과 비용이 많이 든다. 본 연구에서는 위의 문제를 극복하고자 random forest, light gradient boosting machine, logistic regression 모델을 제안한다. 모델은 입력으로 화합물의 분자 구조와 물리화학적 특징을 사용하고 출력으로 약인성 간 손상을 예측한다. 최적의 모델은 평가 지표에서 전반적으로 좋은 성능을 보인 random forest였다. 본 연구에서 제안된 모델은 신약 후보물질의 잠재적인 간 손상을 미리 파악함으로써 신약 개발 과정에 도움을 줄 수 있을 것으로 기대된다.
2022
22.
Sangyun Lee; Soyeon Lee; Myeonghyeon Jeong; Sunwoo Jung; Myoungjin Lee; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Cataracts, Medical informatics, NHANES, Nutrients, Nutrition surveys
@article{lee2022relationship,
title = {The relationship between nutrient intake and cataracts in the older adult population of Korea},
author = {Sangyun Lee and Soyeon Lee and Myeonghyeon Jeong and Sunwoo Jung and Myoungjin Lee and Sunyong Yoo},
url = {https://www.mdpi.com/2072-6643/14/23/4962},
doi = {10.3390/nu14234962},
year = {2022},
date = {2022-11-23},
urldate = {2022-11-23},
journal = {Nutrients},
volume = {14},
number = {23},
pages = {4962},
publisher = {MDPI},
abstract = {Cataracts are a prevalent ophthalmic disease worldwide, and research on the risk factors for cataracts occurrence is actively being conducted. This study aimed to investigate the relationship between nutrient intake and cataracts in the older adult population in Korea. We analyzed data from Korean adults over the age of 60 years (cataract: 2137, non-cataract: 3497) using the Korean National Health and Nutrition Examination Survey. We performed univariate simple and multiple logistic regressions, adjusting for socio-demographic, medical history, and lifestyle, to identify the associations between nutrient intake and cataracts. A higher intake of vitamin B1 in the male group was associated with a lower incidence of cataracts. A lower intake of polyunsaturated fatty acids and vitamin A, and a higher intake of vitamin B2 in the female group were associated with a higher incidence of cataracts. Our study demonstrated that polyunsaturated fatty acids, vitamin A, and vitamin B2 could affect the incidence of cataracts according to sex. The findings could be used to control nutrient intake for cataract prevention.},
note = {Correspondence to Sunyong Yoo},
keywords = {Cataracts, Medical informatics, NHANES, Nutrients, Nutrition surveys},
pubstate = {published},
tppubtype = {article}
}
Cataracts are a prevalent ophthalmic disease worldwide, and research on the risk factors for cataracts occurrence is actively being conducted. This study aimed to investigate the relationship between nutrient intake and cataracts in the older adult population in Korea. We analyzed data from Korean adults over the age of 60 years (cataract: 2137, non-cataract: 3497) using the Korean National Health and Nutrition Examination Survey. We performed univariate simple and multiple logistic regressions, adjusting for socio-demographic, medical history, and lifestyle, to identify the associations between nutrient intake and cataracts. A higher intake of vitamin B1 in the male group was associated with a lower incidence of cataracts. A lower intake of polyunsaturated fatty acids and vitamin A, and a higher intake of vitamin B2 in the female group were associated with a higher incidence of cataracts. Our study demonstrated that polyunsaturated fatty acids, vitamin A, and vitamin B2 could affect the incidence of cataracts according to sex. The findings could be used to control nutrient intake for cataract prevention.
21.
Jin Hyo Park; Su Yeon Kim; Dong Young Kim; Geon Kim; Je Won Park; Sunyong Yoo; Young-Woo Lee; Myoung Jin Lee
Abstract | Links | BibTeX | Dimensions | Tags: Optimization
@article{park2022row,
title = {Row hammer reduction using a buried insulator in a buried channel array transistor},
author = {Jin Hyo Park and Su Yeon Kim and Dong Young Kim and Geon Kim and Je Won Park and Sunyong Yoo and Young-Woo Lee and Myoung Jin Lee},
url = {https://ieeexplore.ieee.org/abstract/document/9938404},
doi = {10.1109/TED.2022.3215931},
year = {2022},
date = {2022-11-03},
urldate = {2022-11-03},
journal = {IEEE Transactions on Electron Devices},
volume = {69},
number = {12},
pages = {6710–6716},
publisher = {IEEE},
abstract = {In this article, we propose an analysis of the usage of a partial isolation type buried channel array transistor (Pi-BCAT). Compared with other structures, the conventional BCAT exhibits improved characteristics in the row hammer effect (RHE) because of its shallow drain/body (D/B) junction. Nevertheless, it remains affected by the RHE and should be mitigated because it is directly related to the reliability of dynamic random access memory (DRAM) applications. The proposed device exhibits a 50% improvement in the RHE and reduces leakage current ( IOFF ) to one-third the level of conventional BCATs while also minimizing the ON -current ( ION ) reduction. Moreover, to efficiently compare RHE, we compare ΔVSN by RHE and ΔVSN based on the gate-induced drain leakage (GIDL) according to bias conditions and the device’s parameters. Finally, we optimize the parameter values of the buried insulator by considering electrical characteristics and the RHE.},
note = {Correspondence to Myoung Jin Lee},
keywords = {Optimization},
pubstate = {published},
tppubtype = {article}
}
In this article, we propose an analysis of the usage of a partial isolation type buried channel array transistor (Pi-BCAT). Compared with other structures, the conventional BCAT exhibits improved characteristics in the row hammer effect (RHE) because of its shallow drain/body (D/B) junction. Nevertheless, it remains affected by the RHE and should be mitigated because it is directly related to the reliability of dynamic random access memory (DRAM) applications. The proposed device exhibits a 50% improvement in the RHE and reduces leakage current ( IOFF ) to one-third the level of conventional BCATs while also minimizing the ON -current ( ION ) reduction. Moreover, to efficiently compare RHE, we compare ΔVSN by RHE and ΔVSN based on the gate-induced drain leakage (GIDL) according to bias conditions and the device’s parameters. Finally, we optimize the parameter values of the buried insulator by considering electrical characteristics and the RHE.
20.
Seonwoo Jung; Min-Keun Song; Eunjoo Lee; Sejin Bae; Yeon-Yong Kim; Doheon Lee; Myoung Jin Lee; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Atrial fibrillation, Attention mechanism, Deep learning, Machine learning, Medical informatics, National health insurance service, Stroke
@article{jung2022predicting,
title = {Predicting ischemic stroke in patients with atrial fibrillation using machine learning},
author = {Seonwoo Jung and Min-Keun Song and Eunjoo Lee and Sejin Bae and Yeon-Yong Kim and Doheon Lee and Myoung Jin Lee and Sunyong Yoo},
url = {https://www.imrpress.com/journal/FBL/27/3/10.31083/j.fbl2703080/htm?utm_source=TrendMD&utm_medium=cpc&utm_campaign=Frontiers_in_Bioscience-Landmark_TrendMD_1},
doi = {10.31083/j.fbl2703080},
year = {2022},
date = {2022-03-04},
urldate = {2022-03-04},
journal = {Frontiers in Bioscience-Landmark},
volume = {27},
number = {3},
pages = {80},
publisher = {IMR Press},
abstract = {Background
Atrial fibrillation (AF) is a well-known risk factor for stroke. Predicting the risk is important to prevent the first and secondary attacks of cerebrovascular diseases by determining early treatment. This study aimed to predict the ischemic stroke in AF patients based on the massive and complex Korean National Health Insurance (KNHIS) data through a machine learning approach.
Methods
We extracted 65-dimensional features, including demographics, health examination, and medical history information, of 754,949 patients with AF from KNHIS. Logistic regression was used to determine whether the extracted features had a statistically significant association with ischemic stroke occurrence. Then, we constructed the ischemic stroke prediction model using an attention-based deep neural network. The extracted features were used as input, and the occurrence of ischemic stroke after the diagnosis of AF was the output used to train the model.
Results We found 48 features significantly associated with ischemic stroke occurrence through regression analysis (p-value < 0.001). When the proposed deep learning model was applied to 150,989 AF patients, it was confirmed that the occurrence ischemic stroke was predicted to be higher AUROC (AUROC = 0.727 ± 0.003) compared to CHA2DS2-VASc score (AUROC = 0.651 ± 0.007) and other machine learning methods.
Conclusions
As part of preventive medicine, this study could help AF patients prepare for ischemic stroke prevention based on predicted stoke associated features and risk scores.},
note = {Correspondence to Sunyong Yoo},
keywords = {Atrial fibrillation, Attention mechanism, Deep learning, Machine learning, Medical informatics, National health insurance service, Stroke},
pubstate = {published},
tppubtype = {article}
}
Background
Atrial fibrillation (AF) is a well-known risk factor for stroke. Predicting the risk is important to prevent the first and secondary attacks of cerebrovascular diseases by determining early treatment. This study aimed to predict the ischemic stroke in AF patients based on the massive and complex Korean National Health Insurance (KNHIS) data through a machine learning approach.
Methods
We extracted 65-dimensional features, including demographics, health examination, and medical history information, of 754,949 patients with AF from KNHIS. Logistic regression was used to determine whether the extracted features had a statistically significant association with ischemic stroke occurrence. Then, we constructed the ischemic stroke prediction model using an attention-based deep neural network. The extracted features were used as input, and the occurrence of ischemic stroke after the diagnosis of AF was the output used to train the model.
Results We found 48 features significantly associated with ischemic stroke occurrence through regression analysis (p-value < 0.001). When the proposed deep learning model was applied to 150,989 AF patients, it was confirmed that the occurrence ischemic stroke was predicted to be higher AUROC (AUROC = 0.727 ± 0.003) compared to CHA2DS2-VASc score (AUROC = 0.651 ± 0.007) and other machine learning methods.
Conclusions
As part of preventive medicine, this study could help AF patients prepare for ischemic stroke prevention based on predicted stoke associated features and risk scores.
Atrial fibrillation (AF) is a well-known risk factor for stroke. Predicting the risk is important to prevent the first and secondary attacks of cerebrovascular diseases by determining early treatment. This study aimed to predict the ischemic stroke in AF patients based on the massive and complex Korean National Health Insurance (KNHIS) data through a machine learning approach.
Methods
We extracted 65-dimensional features, including demographics, health examination, and medical history information, of 754,949 patients with AF from KNHIS. Logistic regression was used to determine whether the extracted features had a statistically significant association with ischemic stroke occurrence. Then, we constructed the ischemic stroke prediction model using an attention-based deep neural network. The extracted features were used as input, and the occurrence of ischemic stroke after the diagnosis of AF was the output used to train the model.
Results We found 48 features significantly associated with ischemic stroke occurrence through regression analysis (p-value < 0.001). When the proposed deep learning model was applied to 150,989 AF patients, it was confirmed that the occurrence ischemic stroke was predicted to be higher AUROC (AUROC = 0.727 ± 0.003) compared to CHA2DS2-VASc score (AUROC = 0.651 ± 0.007) and other machine learning methods.
Conclusions
As part of preventive medicine, this study could help AF patients prepare for ischemic stroke prevention based on predicted stoke associated features and risk scores.
2021
19.
Jinmyung Jung; Yongdeuk Hwang; Hongryul Ahn; Sunjae Lee; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Cancer therapeutics, Genetic interaction, Network analysis, Refining process
@article{Jung2021,
title = {Precise Characterization of Genetic Interactions in Cancer via Molecular Network Refining Processes},
author = {Jinmyung Jung and Yongdeuk Hwang and Hongryul Ahn and Sunjae Lee and Sunyong Yoo},
url = {https://www.mdpi.com/1422-0067/22/20/11114},
doi = {10.3390/ijms222011114},
year = {2021},
date = {2021-10-15},
urldate = {2021-10-15},
journal = {International journal of molecular sciences},
volume = {22},
number = {20},
pages = {11114},
publisher = {MDPI},
abstract = {Genetic interactions (GIs), such as the synthetic lethal interaction, are promising therapeutic targets in precision medicine. However, despite extensive efforts to characterize GIs by large-scale perturbation screening, considerable false positives have been reported in multiple studies. We propose a new computational approach for improved precision in GI identification by applying constraints that consider actual biological phenomena. In this study, GIs were characterized by assessing mutation, loss of function, and expression profiles in the DEPMAP database. The expression profiles were used to exclude loss-of-function data for nonexpressed genes in GI characterization. More importantly, the characterized GIs were refined based on Kyoto Encyclopedia of Genes and Genomes (KEGG) or protein–protein interaction (PPI) networks, under the assumption that genes genetically interacting with a certain mutated gene are adjacent in the networks. As a result, the initial GIs characterized with CRISPR and RNAi screenings were refined to 65 and 23 GIs based on KEGG networks and to 183 and 142 GIs based on PPI networks. The evaluation of refined GIs showed improved precision with respect to known synthetic lethal interactions. The refining process also yielded a synthetic partner network (SPN) for each mutated gene, which provides insight into therapeutic strategies for the mutated genes; specifically, exploring the SPN of mutated BRAF revealed ELAVL1 as a potential target for treating BRAF-mutated cancer, as validated by previous research. We expect that this work will advance cancer therapeutic research.},
note = {Correspondence to Sunyong Yoo},
keywords = {Cancer therapeutics, Genetic interaction, Network analysis, Refining process},
pubstate = {published},
tppubtype = {article}
}
Genetic interactions (GIs), such as the synthetic lethal interaction, are promising therapeutic targets in precision medicine. However, despite extensive efforts to characterize GIs by large-scale perturbation screening, considerable false positives have been reported in multiple studies. We propose a new computational approach for improved precision in GI identification by applying constraints that consider actual biological phenomena. In this study, GIs were characterized by assessing mutation, loss of function, and expression profiles in the DEPMAP database. The expression profiles were used to exclude loss-of-function data for nonexpressed genes in GI characterization. More importantly, the characterized GIs were refined based on Kyoto Encyclopedia of Genes and Genomes (KEGG) or protein–protein interaction (PPI) networks, under the assumption that genes genetically interacting with a certain mutated gene are adjacent in the networks. As a result, the initial GIs characterized with CRISPR and RNAi screenings were refined to 65 and 23 GIs based on KEGG networks and to 183 and 142 GIs based on PPI networks. The evaluation of refined GIs showed improved precision with respect to known synthetic lethal interactions. The refining process also yielded a synthetic partner network (SPN) for each mutated gene, which provides insight into therapeutic strategies for the mutated genes; specifically, exploring the SPN of mutated BRAF revealed ELAVL1 as a potential target for treating BRAF-mutated cancer, as validated by previous research. We expect that this work will advance cancer therapeutic research.
18.
Jin Hyo Park; Geon Kim; Dong Yeong Kim; Su Yeon Kim; Sunyong Yoo; Myoung Jin Lee
Abstract | Links | BibTeX | Dimensions | Tags: Optimization
@article{park2021s,
title = {S-TAT leakage current in partial isolation type saddle-FinFET (Pi-FinFET) s},
author = {Jin Hyo Park and Geon Kim and Dong Yeong Kim and Su Yeon Kim and Sunyong Yoo and Myoung Jin Lee},
url = {https://ieeexplore.ieee.org/abstract/document/9507492},
doi = {10.1109/ACCESS.2021.3102687},
year = {2021},
date = {2021-08-05},
urldate = {2021-08-05},
journal = {IEEE Access},
volume = {9},
pages = {111567–111575},
publisher = {IEEE},
abstract = {In this paper, we compare conventional saddle type FinFETs to partial isolation type saddle FinFETs (Pi-FinFETs) using 3D TCAD simulations to examine the effect of single charge traps for proper prediction of leakage current. We simulated single charge traps at various locations in the drain region, and analyzed how the traps affect leakage current. Our results show that Pi-FinFETs enhanced the leakage current characteristics given the presence of a single charge trap. Also, it was found that Pi-FinFETs exhibit half the FTAT of S-FinFETs. Based on the results from our analysis method, where we use Ioff fluctuation, the FTAT , the σF and the PF parameters to accurately compare performance, and present device design guidelines aimed at improving DRAM refresh characteristics.},
keywords = {Optimization},
pubstate = {published},
tppubtype = {article}
}
In this paper, we compare conventional saddle type FinFETs to partial isolation type saddle FinFETs (Pi-FinFETs) using 3D TCAD simulations to examine the effect of single charge traps for proper prediction of leakage current. We simulated single charge traps at various locations in the drain region, and analyzed how the traps affect leakage current. Our results show that Pi-FinFETs enhanced the leakage current characteristics given the presence of a single charge trap. Also, it was found that Pi-FinFETs exhibit half the FTAT of S-FinFETs. Based on the results from our analysis method, where we use Ioff fluctuation, the FTAT , the σF and the PF parameters to accurately compare performance, and present device design guidelines aimed at improving DRAM refresh characteristics.
17.
Zaki Masood; Hosung Park; Han Seung Jang; Sunyong Yoo; Sokhee P Jung; Yonghoon Choi
Abstract | Links | BibTeX | Dimensions | Tags: Optimization
@article{masood2020optimalc,
title = {Optimal power allocation for maximizing energy efficiency in DAS-based IoT network},
author = {Zaki Masood and Hosung Park and Han Seung Jang and Sunyong Yoo and Sokhee P Jung and Yonghoon Choi},
url = {https://ieeexplore.ieee.org/abstract/document/9166712},
doi = {10.1109/JSYST.2020.3013693},
year = {2021},
date = {2021-06-01},
urldate = {2021-06-01},
journal = {IEEE Systems Journal},
volume = {15},
number = {2},
pages = {2342–2348},
publisher = {IEEE},
abstract = {Distributed antenna system based on simultaneous wireless information and power transfer (SWIPT) can be one of the promising solutions in maximizing energy efficiency (EE), where ultra low power devices harvest energy in power splitting (PS) mode. The paradigm shift of the internet-of-things (IoT) has increased the number of IoT devices and associated sensitive data exchange on the internet. Like the EE is a noteworthy aspect in ultra low power devices, energy harvesting (EH) is an active approach from surrounding electromagnetic sources. This article deals with EE maximization for SWIPT using PS mode. In the SWIPT system, this article presents a tradeoff between EE and spectral efficiency and proposes an algorithm, which allocates optimal power to each distributed antenna port. For an IoT device, the PS scheme implements EH and information decoding operations. The proposed algorithm is based on the Lagrangian multiplier method and Karush-Kuhn-Tucker conditions to find the optimal solution without iterative computation compared to the conventional iterative method. Simulation results reveal that the proposed algorithm achieves maximum energy transfer by the using optimal PS ratio.},
keywords = {Optimization},
pubstate = {published},
tppubtype = {article}
}
Distributed antenna system based on simultaneous wireless information and power transfer (SWIPT) can be one of the promising solutions in maximizing energy efficiency (EE), where ultra low power devices harvest energy in power splitting (PS) mode. The paradigm shift of the internet-of-things (IoT) has increased the number of IoT devices and associated sensitive data exchange on the internet. Like the EE is a noteworthy aspect in ultra low power devices, energy harvesting (EH) is an active approach from surrounding electromagnetic sources. This article deals with EE maximization for SWIPT using PS mode. In the SWIPT system, this article presents a tradeoff between EE and spectral efficiency and proposes an algorithm, which allocates optimal power to each distributed antenna port. For an IoT device, the PS scheme implements EH and information decoding operations. The proposed algorithm is based on the Lagrangian multiplier method and Karush-Kuhn-Tucker conditions to find the optimal solution without iterative computation compared to the conventional iterative method. Simulation results reveal that the proposed algorithm achieves maximum energy transfer by the using optimal PS ratio.
16.
Hyeonseo Yun; Dong-Wook Kim; Eun-Joo Lee; Jinmyung Jung; Sunyong Yoo
Abstract | Links | BibTeX | Dimensions | Tags: Depression, Dietary habits, Medical informatics, NHANES, Nutrients, Nutrition surveys
@article{yun2021analysis,
title = {Analysis of the effects of nutrient intake and dietary habits on depression in Korean adults},
author = {Hyeonseo Yun and Dong-Wook Kim and Eun-Joo Lee and Jinmyung Jung and Sunyong Yoo},
url = {https://www.mdpi.com/2072-6643/13/4/1360},
doi = {10.3390/nu13041360},
year = {2021},
date = {2021-04-19},
urldate = {2021-04-19},
journal = {Nutrients},
volume = {13},
number = {4},
pages = {1360},
publisher = {MDPI},
abstract = {While several studies have explored nutrient intake and dietary habits associated with depression, few studies have reflected recent trends and demographic factors. Therefore, we examined how nutrient intake and eating habits are associated with depression, according to gender and age. We performed simple and multiple regressions using nationally representative samples of 10,106 subjects from the Korea National Health and Nutrition Examination Survey. The results indicated that cholesterol, dietary fiber, sodium, frequency of breakfast, lunch, dinner, and eating out were significantly associated with depression (p-value < 0.05). Moreover, depression was associated with nutrient intake and dietary habits by gender and age group: sugar, breakfast, lunch, and eating out frequency in the young women’s group; sodium and lunch frequency among middle-age men; dietary fibers, breakfast, and eating out frequency among middle-age women; energy, moisture, carbohydrate, lunch, and dinner frequency in late middle-age men; breakfast and lunch frequency among late middle-age women; vitamin A, carotene, lunch, and eating out frequency among older age men; and fat, saturated fatty acids, omega-3 fatty acid, omega-6 fatty acid, and eating out frequency among the older age women’s group (p-value < 0.05). This study can be used to establish dietary strategies for depression prevention, considering gender and age.},
note = {Correspondence to Sunyong Yoo},
keywords = {Depression, Dietary habits, Medical informatics, NHANES, Nutrients, Nutrition surveys},
pubstate = {published},
tppubtype = {article}
}
While several studies have explored nutrient intake and dietary habits associated with depression, few studies have reflected recent trends and demographic factors. Therefore, we examined how nutrient intake and eating habits are associated with depression, according to gender and age. We performed simple and multiple regressions using nationally representative samples of 10,106 subjects from the Korea National Health and Nutrition Examination Survey. The results indicated that cholesterol, dietary fiber, sodium, frequency of breakfast, lunch, dinner, and eating out were significantly associated with depression (p-value < 0.05). Moreover, depression was associated with nutrient intake and dietary habits by gender and age group: sugar, breakfast, lunch, and eating out frequency in the young women’s group; sodium and lunch frequency among middle-age men; dietary fibers, breakfast, and eating out frequency among middle-age women; energy, moisture, carbohydrate, lunch, and dinner frequency in late middle-age men; breakfast and lunch frequency among late middle-age women; vitamin A, carotene, lunch, and eating out frequency among older age men; and fat, saturated fatty acids, omega-3 fatty acid, omega-6 fatty acid, and eating out frequency among the older age women’s group (p-value < 0.05). This study can be used to establish dietary strategies for depression prevention, considering gender and age.
15.
Kiseong Kim; Sunyong Yoo; Sangyeon Lee; Doheon Lee; Kwang-Hyung Lee
Abstract | Links | BibTeX | Dimensions | Tags: Disease spread, Epidemic disease, Network analysis, Pandemic
@article{kim2021network,
title = {Network analysis to identify the risk of epidemic spreading},
author = {Kiseong Kim and Sunyong Yoo and Sangyeon Lee and Doheon Lee and Kwang-Hyung Lee},
url = {https://www.mdpi.com/2076-3417/11/7/2997},
doi = {10.3390/app11072997},
year = {2021},
date = {2021-03-26},
urldate = {2021-03-26},
journal = {Applied Sciences},
volume = {11},
number = {7},
pages = {2997},
publisher = {MDPI},
abstract = {Several epidemics, such as the Black Death and the Spanish flu, have threatened human life throughout history; however, it is unclear if humans will remain safe from the sudden and fast spread of epidemic diseases. Moreover, the transmission characteristics of epidemics remain undiscovered. In this study, we present the results of an epidemic simulation experiment revealing the relationship between epidemic parameters and pandemic risk. To analyze the time-dependent risk and impact of epidemics, we considered two parameters for infectious diseases: the recovery time from infection and the transmission rate of the disease. Based on the epidemic simulation, we identified two important aspects of human safety with regard to the threat of a pandemic. First, humans should be safe if the fatality rate is below 100%. Second, even when the fatality rate is 100%, humans would be safe if the average degree of human social networks is below a threshold value. Nevertheless, certain diseases can potentially infect all nodes in the human social networks, and these diseases cause a pandemic when the average degree is larger than the threshold value. These results indicated that certain infectious diseases lead to human extinction and can be prevented by minimizing human contact.},
keywords = {Disease spread, Epidemic disease, Network analysis, Pandemic},
pubstate = {published},
tppubtype = {article}
}
Several epidemics, such as the Black Death and the Spanish flu, have threatened human life throughout history; however, it is unclear if humans will remain safe from the sudden and fast spread of epidemic diseases. Moreover, the transmission characteristics of epidemics remain undiscovered. In this study, we present the results of an epidemic simulation experiment revealing the relationship between epidemic parameters and pandemic risk. To analyze the time-dependent risk and impact of epidemics, we considered two parameters for infectious diseases: the recovery time from infection and the transmission rate of the disease. Based on the epidemic simulation, we identified two important aspects of human safety with regard to the threat of a pandemic. First, humans should be safe if the fatality rate is below 100%. Second, even when the fatality rate is 100%, humans would be safe if the average degree of human social networks is below a threshold value. Nevertheless, certain diseases can potentially infect all nodes in the human social networks, and these diseases cause a pandemic when the average degree is larger than the threshold value. These results indicated that certain infectious diseases lead to human extinction and can be prevented by minimizing human contact.
14.
Sunyong Yoo; Dong-Wook Kim; Young-Eun Kim; Jong Heon Park; Yeon-Yong Kim; Kyu-dong Cho; Mi-Ji Gwon; Jae-In Shin; Eun-Joo Lee
Abstract | Links | BibTeX | Dimensions | Tags: Allergic rhinitis, Asthma, Atopic dermatitis, Database, National health insurance service
@article{yoo2021data,
title = {Data resource profile: the allergic disease database of the Korean National Health Insurance Service},
author = {Sunyong Yoo and Dong-Wook Kim and Young-Eun Kim and Jong Heon Park and Yeon-Yong Kim and Kyu-dong Cho and Mi-Ji Gwon and Jae-In Shin and Eun-Joo Lee},
url = {https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8060521/},
doi = {10.4178/epih.e2021010},
year = {2021},
date = {2021-01-21},
urldate = {2021-01-21},
journal = {Epidemiology and Health},
volume = {43},
pages = {e2021010},
publisher = {Korean Society of Epidemiology},
abstract = {Researchers have been interested in probing how the environmental factors associated with allergic diseases affect the use of medical services. Considering this demand, we have constructed a database, named the Allergic Disease Database, based on the National Health Insurance Database (NHID). The NHID contains information on demographic and medical service utilization for approximately 99% of the Korean population. This study targeted 3 major allergic diseases, including allergic rhinitis, atopic dermatitis, and asthma. For the target diseases, our database provides daily medical service information, including the number of daily visits from 2013 and 2017, categorized by patients’ characteristics such as address, sex, age, and duration of residence. We provide additional information, including yearly population, a number of patients, and averaged geocoding coordinates by eup, myeon, and dong district code (the smallest-scale administrative units in Korea). This information enables researchers to analyze how daily changes in the environmental factors of allergic diseases (e.g., particulate matter, sulfur dioxide, and ozone) in certain regions would influence patients’ behavioral patterns of medical service utilization. Moreover, researchers can analyze long-term trends in allergic diseases and the health effects caused by environmental factors such as daily climate and pollution data. The advantages of this database are easy access to data, additional levels of geographic detail, time-efficient data-refining and processing, and a de-identification process that minimizes the exposure of identifiable personal information. All datasets included in the Allergic Disease Database can be downloaded by accessing the National Health Insurance Service data sharing webpage (https://nhiss.nhis.or.kr).},
keywords = {Allergic rhinitis, Asthma, Atopic dermatitis, Database, National health insurance service},
pubstate = {published},
tppubtype = {article}
}
Researchers have been interested in probing how the environmental factors associated with allergic diseases affect the use of medical services. Considering this demand, we have constructed a database, named the Allergic Disease Database, based on the National Health Insurance Database (NHID). The NHID contains information on demographic and medical service utilization for approximately 99% of the Korean population. This study targeted 3 major allergic diseases, including allergic rhinitis, atopic dermatitis, and asthma. For the target diseases, our database provides daily medical service information, including the number of daily visits from 2013 and 2017, categorized by patients’ characteristics such as address, sex, age, and duration of residence. We provide additional information, including yearly population, a number of patients, and averaged geocoding coordinates by eup, myeon, and dong district code (the smallest-scale administrative units in Korea). This information enables researchers to analyze how daily changes in the environmental factors of allergic diseases (e.g., particulate matter, sulfur dioxide, and ozone) in certain regions would influence patients’ behavioral patterns of medical service utilization. Moreover, researchers can analyze long-term trends in allergic diseases and the health effects caused by environmental factors such as daily climate and pollution data. The advantages of this database are easy access to data, additional levels of geographic detail, time-efficient data-refining and processing, and a de-identification process that minimizes the exposure of identifiable personal information. All datasets included in the Allergic Disease Database can be downloaded by accessing the National Health Insurance Service data sharing webpage (https://nhiss.nhis.or.kr).
13.
윤현서; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Transformer
@article{윤현서2021transformer,
title = {Transformer 기반 비윤리적 문장 탐지},
author = {윤현서 and 유선용},
url = {https://www.dbpia.co.kr/pdf/pdfView.do?nodeId=NODE10595980&googleIPSandBox=false&mark=0&minRead=5&ipRange=false&b2cLoginYN=false&icstClss=010000&isPDFSizeAllowed=true&accessgl=Y&language=ko_KR&hasTopBanner=true},
doi = {10.9728/dcs.2021},
year = {2021},
date = {2021-01-01},
urldate = {2021-01-01},
journal = {디지털콘텐츠학회논문지},
volume = {22},
number = {8},
pages = {1289–1293},
abstract = {정보통신 기술의 발달은 사회관계망서비스(SNS)의 확산을 가져왔지만 심각한 사회적 문제인 악성 댓글을 야기하였다. 사이버 명예훼손ᆞ모욕 발생/검거 건수는 2014년 8,880건에서 2019년 16,633건으로 급격히 증가하였고 해당 문제를 해결하기 위한 대책이 요구된다. 그러나 IP 블랙리스트, 비속어 필터와 같은 기존의 규제만으로는 다양한 패턴을 가지는 악성 댓글을 탐지하는데 한계가 있다. 따라서 비윤리적 문장 탐지에 최적화된 인공지능 모델이 필요하다. 본 논문은 자연어 처리에서 높은 성능을 보여준 Transformer 기반 비윤리적 문장 탐지 모델을 제안한다. 해당 모델은 95.03%의 정확도를 보여주었고 비윤리적 문장 탐지 모델로 활용될 것이다. 또한, SNS의 댓글뿐만 아니라 스트리밍 서비스 등 다양한 분야에},
keywords = {Transformer},
pubstate = {published},
tppubtype = {article}
}
정보통신 기술의 발달은 사회관계망서비스(SNS)의 확산을 가져왔지만 심각한 사회적 문제인 악성 댓글을 야기하였다. 사이버 명예훼손ᆞ모욕 발생/검거 건수는 2014년 8,880건에서 2019년 16,633건으로 급격히 증가하였고 해당 문제를 해결하기 위한 대책이 요구된다. 그러나 IP 블랙리스트, 비속어 필터와 같은 기존의 규제만으로는 다양한 패턴을 가지는 악성 댓글을 탐지하는데 한계가 있다. 따라서 비윤리적 문장 탐지에 최적화된 인공지능 모델이 필요하다. 본 논문은 자연어 처리에서 높은 성능을 보여준 Transformer 기반 비윤리적 문장 탐지 모델을 제안한다. 해당 모델은 95.03%의 정확도를 보여주었고 비윤리적 문장 탐지 모델로 활용될 것이다. 또한, SNS의 댓글뿐만 아니라 스트리밍 서비스 등 다양한 분야에
12.
이소연; 최지은; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Attention mechanism
@article{이소연2021attention,
title = {Attention 알고리즘 기반 요약 콘텐츠 생성 방안 연구},
author = {이소연 and 최지은 and 유선용},
url = {http://journal.dcs.or.kr/_PR/view/?aidx=30553&bidx=2701},
doi = {10.9728/dcs.2021.22.9.1487},
year = {2021},
date = {2021-01-01},
urldate = {2021-01-01},
journal = {Journal of Digital Contents Society},
volume = {22},
number = {9},
pages = {1487–1491},
abstract = {최근 바쁜 현대인들에게 뉴스, 도서, 영화, TV 프로그램 등 각종 콘텐츠를 요약해 제공하는 ‘요약 콘텐츠(Summary Contents)’ 시장이 주목받고 있다. 기존 대부분의 콘텐츠 요약 기법은 문장을 분석하여 통계적으로 의미있는 단어를 추출하는 것에 집중하였다. 하지만 단순히 단어의 구문적 특징만을 고려할 경우 단어들 간의 연관성과 내재된 의미를 놓치는 경우가 많다. 따라서, 문장의 복잡한 구조와 의미를 고려하여 핵심 요소를 추출하고 추상적 요약을 만들기 위한 방법이 필요하다. 본 연구는 영문 리뷰 데이터와 국문 신문 기사 데이터에 attention 알고리즘 기반 딥러닝 모델을 적용하여 핵심 문맥을 반영한 추상적 요약문을 생성한다. 실험 결과, 제안하는 모델은 단어의 의미를 중점적으로 해석해 성공적으로 영문 리뷰 데이터의 요약 예측문을 생성하였다. 국문 텍스트의 경우 전처리가 까다로움에도 실제와 유사한 예측 요약문을 생성하는 유의미한 결과를 보였다. 수기 확인(manual curation) 및 설문조사 결과, 생성된 요약 콘텐츠는 주요 단어 및 추상적 개념을 효과적으로 생성하여 문장을 요약하는 것을 확인할 수 있었다. 본 연구는 향후 현대인들에게 정보를 전달하는 과정에서 시간 단축 및 편리성을 제공할 수 있을 것이다.},
keywords = {Attention mechanism},
pubstate = {published},
tppubtype = {article}
}
최근 바쁜 현대인들에게 뉴스, 도서, 영화, TV 프로그램 등 각종 콘텐츠를 요약해 제공하는 ‘요약 콘텐츠(Summary Contents)’ 시장이 주목받고 있다. 기존 대부분의 콘텐츠 요약 기법은 문장을 분석하여 통계적으로 의미있는 단어를 추출하는 것에 집중하였다. 하지만 단순히 단어의 구문적 특징만을 고려할 경우 단어들 간의 연관성과 내재된 의미를 놓치는 경우가 많다. 따라서, 문장의 복잡한 구조와 의미를 고려하여 핵심 요소를 추출하고 추상적 요약을 만들기 위한 방법이 필요하다. 본 연구는 영문 리뷰 데이터와 국문 신문 기사 데이터에 attention 알고리즘 기반 딥러닝 모델을 적용하여 핵심 문맥을 반영한 추상적 요약문을 생성한다. 실험 결과, 제안하는 모델은 단어의 의미를 중점적으로 해석해 성공적으로 영문 리뷰 데이터의 요약 예측문을 생성하였다. 국문 텍스트의 경우 전처리가 까다로움에도 실제와 유사한 예측 요약문을 생성하는 유의미한 결과를 보였다. 수기 확인(manual curation) 및 설문조사 결과, 생성된 요약 콘텐츠는 주요 단어 및 추상적 개념을 효과적으로 생성하여 문장을 요약하는 것을 확인할 수 있었다. 본 연구는 향후 현대인들에게 정보를 전달하는 과정에서 시간 단축 및 편리성을 제공할 수 있을 것이다.
11.
정선우; 이민지; 유선용
Abstract | Links | BibTeX | Dimensions | Tags: Machine learning, Medical informatics
@article{정선우2021공공빅데이터를,
title = {공공빅데이터를 활용한 기계학습 기반 뇌졸중 위험도 예측},
author = {정선우 and 이민지 and 유선용},
url = {https://kiss.kstudy.com/Detail/Ar?key=3863715},
doi = {10.12673/jant.2021.25.1.96},
year = {2021},
date = {2021-01-01},
urldate = {2021-01-01},
journal = {한국항행학회논문지},
volume = {25},
number = {1},
pages = {96–101},
publisher = {한국항행학회},
abstract = {본 논문은 빅데이터를 이용하여 심방세동 환자의 뇌졸중 발병을 예측하는 기계 학습 모델을 제시한다. 학습 데이터로는 국민 건강 보험공단에서 제공하는 대한민국 전수에 해당하는 심방세동 환자의 정보를 수집하였다. 수집된 정보는 인구사회학, 과거 병력, 건강검진을 포함한 68개 독립변수로 구성된다. 본 연구의 목표는 기존 심방세동 환자의 뇌졸중 위험도 예측에 사용되던 통계적 모델 (CHADS2, CHA2DS2-VASc)의 성능을 검증하고 기계 학습 모델을 적용하여 기존 모델보다 높은 정확도를 가지는 모델을 제시하는 것이다. 제안하는 모델의 정확도, AUROC (area under the receiver operating characteristic)를 검증한 결과 제안하는 기계 학습 기반의 모형이 심방세동 환자의 뇌졸중 위험도를 사용한 모델이 기존의 통계적 모델보다 높은 정확도, 민감도, 특이도를 가지는 것을 확인할 수 있었다.},
keywords = {Machine learning, Medical informatics},
pubstate = {published},
tppubtype = {article}
}
본 논문은 빅데이터를 이용하여 심방세동 환자의 뇌졸중 발병을 예측하는 기계 학습 모델을 제시한다. 학습 데이터로는 국민 건강 보험공단에서 제공하는 대한민국 전수에 해당하는 심방세동 환자의 정보를 수집하였다. 수집된 정보는 인구사회학, 과거 병력, 건강검진을 포함한 68개 독립변수로 구성된다. 본 연구의 목표는 기존 심방세동 환자의 뇌졸중 위험도 예측에 사용되던 통계적 모델 (CHADS2, CHA2DS2-VASc)의 성능을 검증하고 기계 학습 모델을 적용하여 기존 모델보다 높은 정확도를 가지는 모델을 제시하는 것이다. 제안하는 모델의 정확도, AUROC (area under the receiver operating characteristic)를 검증한 결과 제안하는 기계 학습 기반의 모형이 심방세동 환자의 뇌졸중 위험도를 사용한 모델이 기존의 통계적 모델보다 높은 정확도, 민감도, 특이도를 가지는 것을 확인할 수 있었다.
2020
10.
Junseok Park; Seongkuk Park; Kwangmin Kim; Woochang Hwang; Sunyong Yoo; Gwan-su Yi; Doheon Lee
Abstract | Links | BibTeX | Dimensions | Tags: Clinical trial, Medical informatics
@article{park2020interactive,
title = {An interactive retrieval system for clinical trial studies with context-dependent protocol elements},
author = {Junseok Park and Seongkuk Park and Kwangmin Kim and Woochang Hwang and Sunyong Yoo and Gwan-su Yi and Doheon Lee},
url = {https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0238290},
doi = {10.1371/journal.pone.0238290},
year = {2020},
date = {2020-09-18},
urldate = {2020-09-18},
journal = {PloS one},
volume = {15},
number = {9},
pages = {e0238290},
publisher = {Public Library of Science San Francisco, CA USA},
abstract = {A well-defined protocol for a clinical trial guarantees a successful outcome report. When designing the protocol, most researchers refer to electronic databases and extract protocol elements using a keyword search. However, state-of-the-art database systems only offer text-based searches for user-entered keywords. In this study, we present a database system with a context-dependent and protocol-element-selection function for successfully designing a clinical trial protocol. To do this, we first introduce a database for a protocol retrieval system constructed from individual protocol data extracted from 184,634 clinical trials and 13,210 frame structures of clinical trial protocols. The database contains a variety of semantic information that allows the filtering of protocols during the search operation. Based on the database, we developed a web application called the clinical trial protocol database system (CLIPS; available at https://corus.kaist.edu/clips). This system enables an interactive search by utilizing protocol elements. To enable an interactive search for combinations of protocol elements, CLIPS provides optional next element selection according to the previous element in the form of a connected tree. The validation results show that our method achieves better performance than that of existing databases in predicting phenotypic features.},
keywords = {Clinical trial, Medical informatics},
pubstate = {published},
tppubtype = {article}
}
A well-defined protocol for a clinical trial guarantees a successful outcome report. When designing the protocol, most researchers refer to electronic databases and extract protocol elements using a keyword search. However, state-of-the-art database systems only offer text-based searches for user-entered keywords. In this study, we present a database system with a context-dependent and protocol-element-selection function for successfully designing a clinical trial protocol. To do this, we first introduce a database for a protocol retrieval system constructed from individual protocol data extracted from 184,634 clinical trials and 13,210 frame structures of clinical trial protocols. The database contains a variety of semantic information that allows the filtering of protocols during the search operation. Based on the database, we developed a web application called the clinical trial protocol database system (CLIPS; available at https://corus.kaist.edu/clips). This system enables an interactive search by utilizing protocol elements. To enable an interactive search for combinations of protocol elements, CLIPS provides optional next element selection according to the previous element in the form of a connected tree. The validation results show that our method achieves better performance than that of existing databases in predicting phenotypic features.
9.
Junseok Park; Seongkuk Park; Gwangmin Kim; Kwangmin Kim; Jaegyun Jung; Sunyong Yoo; Gwan-Su Yi; Doheon Lee
Abstract | Links | BibTeX | Dimensions | Tags: Clinical trial, Medical informatics
@article{park2020reliable,
title = {Reliable data collection in participatory trials to assess digital healthcare applications},
author = {Junseok Park and Seongkuk Park and Gwangmin Kim and Kwangmin Kim and Jaegyun Jung and Sunyong Yoo and Gwan-Su Yi and Doheon Lee},
url = {https://ieeexplore.ieee.org/abstract/document/9054970},
doi = {10.1109/ACCESS.2020.2985122},
year = {2020},
date = {2020-01-01},
urldate = {2020-01-01},
journal = {IEEE Access},
volume = {8},
pages = {79472–79490},
publisher = {IEEE},
abstract = {The number of digital healthcare mobile applications in the market is exponentially increasing owing to the development of mobile networks and widespread usage of smartphones. However, only few of these applications have been adequately validated. Like many mobile applications, in general, the use of healthcare applications is considered safe; thus, developers and end users can easily exchange them in the marketplace. However, existing platforms are unsuitable for collecting reliable data for evaluating the effectiveness of the applications. Moreover, these platforms reflect only the perspectives of developers and experts, and not of end users. For instance, typical clinical trial data collection methods are not appropriate for participant-driven assessment of healthcare applications because of their complexity and high cost. Thus, we identified the need for a participant-driven data collection platform for end users that is interpretable, systematic, and sustainable, as a first step to validate the effectiveness of the applications. To collect reliable data in the participatory trial format, we defined distinct stages for data preparation, storage, and sharing. The interpretable data preparation consists of a protocol database system and semantic feature retrieval method that allow a person without professional knowledge to create a protocol. The systematic data storage stage includes calculation of the collected data reliability weight. For sustainable data collection, we integrated a weight method and a future reward distribution function. We validated the methods through statistical tests involving 718 human participants. The results of a validation experiment demonstrate that the compared methods differ significantly and prove that the choice of an appropriate method is essential for reliable data collection, to facilitate effectiveness validation of digital healthcare applications. Furthermore, we created a Web-based system for our pilot platform to collect reliable data in an integrated pipeline. We compared the platform features using existing clinical and pragmatic trial data collection platforms.},
keywords = {Clinical trial, Medical informatics},
pubstate = {published},
tppubtype = {article}
}
The number of digital healthcare mobile applications in the market is exponentially increasing owing to the development of mobile networks and widespread usage of smartphones. However, only few of these applications have been adequately validated. Like many mobile applications, in general, the use of healthcare applications is considered safe; thus, developers and end users can easily exchange them in the marketplace. However, existing platforms are unsuitable for collecting reliable data for evaluating the effectiveness of the applications. Moreover, these platforms reflect only the perspectives of developers and experts, and not of end users. For instance, typical clinical trial data collection methods are not appropriate for participant-driven assessment of healthcare applications because of their complexity and high cost. Thus, we identified the need for a participant-driven data collection platform for end users that is interpretable, systematic, and sustainable, as a first step to validate the effectiveness of the applications. To collect reliable data in the participatory trial format, we defined distinct stages for data preparation, storage, and sharing. The interpretable data preparation consists of a protocol database system and semantic feature retrieval method that allow a person without professional knowledge to create a protocol. The systematic data storage stage includes calculation of the collected data reliability weight. For sustainable data collection, we integrated a weight method and a future reward distribution function. We validated the methods through statistical tests involving 718 human participants. The results of a validation experiment demonstrate that the compared methods differ significantly and prove that the choice of an appropriate method is essential for reliable data collection, to facilitate effectiveness validation of digital healthcare applications. Furthermore, we created a Web-based system for our pilot platform to collect reliable data in an integrated pipeline. We compared the platform features using existing clinical and pragmatic trial data collection platforms.
8.
Sunyong Yoo; Hyung Chae Yang; Seongyeong Lee; Jaewook Shin; Seyoung Min; Eunjoo Lee; Minkeun Song; Doheon Lee
Abstract | Links | BibTeX | Dimensions | Tags: Bioinformatics, Chemical property, Deep learning, Molecular interaction, Natural product, Network analysis, Text mining
@article{10.3389/fphar.2020.584875,
title = {A Deep Learning-Based Approach for Identifying the Medicinal Uses of Plant-Derived Natural Compounds},
author = {Sunyong Yoo and Hyung Chae Yang and Seongyeong Lee and Jaewook Shin and Seyoung Min and Eunjoo Lee and Minkeun Song and Doheon Lee},
url = {https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2020.584875},
doi = {10.3389/fphar.2020.584875},
issn = {1663-9812},
year = {2020},
date = {2020-01-01},
urldate = {2020-01-01},
journal = {Frontiers in Pharmacology},
volume = {11},
pages = {584875},
abstract = {Medicinal plants and their extracts have been used as important sources for drug discovery. In particular, plant-derived natural compounds, including phytochemicals, antioxidants, vitamins, and minerals, are gaining attention as they promote health and prevent disease. Although several in vitro methods have been developed to confirm the biological activities of natural compounds, there is still considerable room to reduce time and cost. To overcome these limitations, several in silico methods have been proposed for conducting large-scale analysis, but they are still limited in terms of dealing with incomplete and heterogeneous natural compound data. Here, we propose a deep learning-based approach to identify the medicinal uses of natural compounds by exploiting massive and heterogeneous drug and natural compound data. The rationale behind this approach is that deep learning can effectively utilize heterogeneous features to alleviate incomplete information. Based on latent knowledge, molecular interactions, and chemical property features, we generated 686 dimensional features for 4,507 natural compounds and 2,882 approved and investigational drugs. The deep learning model was trained using the generated features and verified drug indication information. When the features of natural compounds were applied as input to the trained model, potential efficacies were successfully predicted with high accuracy, sensitivity, and specificity.},
keywords = {Bioinformatics, Chemical property, Deep learning, Molecular interaction, Natural product, Network analysis, Text mining},
pubstate = {published},
tppubtype = {article}
}
Medicinal plants and their extracts have been used as important sources for drug discovery. In particular, plant-derived natural compounds, including phytochemicals, antioxidants, vitamins, and minerals, are gaining attention as they promote health and prevent disease. Although several in vitro methods have been developed to confirm the biological activities of natural compounds, there is still considerable room to reduce time and cost. To overcome these limitations, several in silico methods have been proposed for conducting large-scale analysis, but they are still limited in terms of dealing with incomplete and heterogeneous natural compound data. Here, we propose a deep learning-based approach to identify the medicinal uses of natural compounds by exploiting massive and heterogeneous drug and natural compound data. The rationale behind this approach is that deep learning can effectively utilize heterogeneous features to alleviate incomplete information. Based on latent knowledge, molecular interactions, and chemical property features, we generated 686 dimensional features for 4,507 natural compounds and 2,882 approved and investigational drugs. The deep learning model was trained using the generated features and verified drug indication information. When the features of natural compounds were applied as input to the trained model, potential efficacies were successfully predicted with high accuracy, sensitivity, and specificity.