Article
Biochemical Research Methods
Tianyuan Liu, Bohao Zou, Manman He, Yongfei Hu, Yiying Dou, Tianyu Cui, Puwen Tan, Shaobin Li, Shuan Rao, Yan Huang, Sixi Liu, Kaican Cai, Dong Wang
Summary: This study developed a deep-learning model called LncReader, which uses a multi-head self-attention mechanism to identify dual functional lncRNAs. By comparing it with various machine learning methods, it was demonstrated that LncReader has advantages in predicting dual functional lncRNAs.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Tianyuan Liu, Bohao Zou, Manman He, Yongfei Hu, Yiying Dou, Tianyu Cui, Puwen Tan, Shaobin Li, Shuan Rao, Yan Huang, Sixi Liu, Kaican Cai, Dong Wang
Summary: Recently, we developed a deep-learning model called LncReader with a multi-head self-attention mechanism to identify dual functional long noncoding ribonucleic acids (lncRNAs). Our data showed that LncReader exhibited several advantages over various classical machine learning methods using benchmark datasets. Additionally, independent in-house datasets generated through mass spectrometry proteomics combined with RNA-seq and Ribo-seq confirmed that LncReader achieved the best performance compared to other tools. Therefore, LncReader provides an accurate and practical tool for fast identification of dual functional lncRNAs.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemistry & Molecular Biology
Adnan Adnan, Hongya Wang, Farman Ali, Majdi Khalid, Omar Alghushairy, Raed Alsini
Summary: piwiRNA is a type of ncRNA that cannot produce proteins and plays a crucial role in gamete generation and gene expression regulation. It has various functions such as deadenylation, animal fertility, transposon silencing, anti-viral activity, and endogenous gene regulation. The prediction of piwiRNA is of great significance and improving its prediction is highly desirable.
JOURNAL OF BIOMOLECULAR STRUCTURE & DYNAMICS
(2023)
Article
Biochemical Research Methods
Dalwinder Singh, Akansha Madhawan, Joy Roy
Summary: The study introduces a method based on RNA classifier that can effectively identify and classify different types of RNAs, achieving high accuracy in both multi-class and binary classification problems.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Min Li, Baoying Zhao, Rui Yin, Chengqian Lu, Fei Guo, Min Zeng
Summary: This paper proposes a graph convolutional network-based deep learning model, GraphLncLoc, for predicting the subcellular localization of long non-coding RNAs (lncRNAs). Unlike previous methods, GraphLncLoc transforms lncRNA sequences into de Bruijn graphs, converting the sequence classification problem into a graph classification problem. Experimental results show that GraphLncLoc outperforms traditional machine learning models and existing predictors. Additionally, analysis demonstrates that transforming sequences into graphs yields more distinguishable features and greater robustness. A case study reveals that GraphLncLoc can uncover important motifs for nucleus subcellular localization.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Min Li, Baoying Zhao, Rui Yin, Chengqian Lu, Fei Guo, Min Zeng
Summary: In this study, we proposed GraphLncLoc, a graph convolutional network-based deep learning model, for predicting the subcellular localization of long non-coding RNAs (lncRNAs). By transforming lncRNA sequences into de Bruijn graphs and utilizing graph convolutional networks to learn latent representations, GraphLncLoc achieved better performance and revealed the advantages of sequence-to-graph transformation.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Oncology
Mohammad Amin Kerachian, Marjan Azghandi
Summary: This article introduces a novel method for discovering lncRNAs based on their intergenic location and the methylation level of a single CpG site, which represents epigenetic characteristics.
CANCER CELL INTERNATIONAL
(2022)
Article
Genetics & Heredity
Yunhe Liu, Qiqing Fu, Xueqing Peng, Chaoyu Zhu, Gang Liu, Lei Liu
Summary: The study introduced an attention-based multi-instance learning network architecture for circRNA identification, outperforming existing models. The effectiveness of the attention mechanism was validated on handwritten digit dataset to obtain key sequence loci for circRNA recognition, followed by motif enrichment analysis for circRNA formation.
Article
Multidisciplinary Sciences
Abu Zahid Bin Aziz, Md. Al Mehedi Hasan, Jungpil Shin
Summary: In this paper, a multi-channel convolution neural network using binary encoding was proposed for identifying pseudouridine sites. Tuning hyperparameters with k-fold cross-validation and grid search, promising results were obtained in independent datasets. This method proved to be effective for identifying pseudouridine sites and has been implemented as an easily accessible web server.
Article
Biochemical Research Methods
Dongqing Zhao, Chunqing Wang, Shuai Yan, Ruibing Chen
Summary: Long non-coding RNAs (lncRNAs) play crucial regulatory roles in gene regulation, chromatin remodeling, and development. Understanding the protein interactions of lncRNAs is important for elucidating their functions and mechanisms. Mass spectrometry and other methods have identified many proteins that interact with lncRNAs.
ANALYTICAL BIOCHEMISTRY
(2022)
Article
Computer Science, Information Systems
Ying Wang, Pengfei Zhao, Hongkai Du, Yingxin Cao, Qinke Peng, Laiyi Fu
Summary: Long non-coding RNAs (LncRNAs) play a crucial role in gene expression regulation and other biological processes. Differentiating lncRNAs from protein-coding transcripts helps researchers understand lncRNA formation mechanisms and its regulation in various diseases. Previous methods for identifying lncRNAs, such as bio-sequencing and machine learning, are not always satisfactory due to the complex feature extraction procedures and artifacts in bio-sequencing. Therefore, the proposed deep learning-based framework, lncDLSM, provides a helpful tool for accurately identifying lncRNAs without prior biological knowledge, achieving satisfactory results across different species through transfer learning.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2023)
Article
Genetics & Heredity
Rattaphon Lin, Duangdao Wichadakul
Summary: The tool Xlnc1DCNN can accurately distinguish between long non-coding RNAs and protein-coding transcripts using a one-dimensional convolutional neural network, and the explanation results reveal the feature differences and potential inconsistencies in database annotations.
FRONTIERS IN GENETICS
(2022)
Article
Multidisciplinary Sciences
Hongwei Yan, Qi Liu, Jieming Jiang, Xufang Shen, Lei Zhang, Zhen Yuan, Yumeng Wu, Ying Liu
Summary: This study utilized high-throughput sequencing to analyze the transcriptome of tiger pufferfish gonads, identifying numerous miRNAs and lncRNAs potentially involved in regulating sex-related gene expression. Differential expression analysis revealed specific miRNAs and lncRNAs that were up-regulated or down-regulated in male gonads compared to female gonads. These findings could lead to further research on the function of miRNAs and lncRNAs in sex determination and differentiation.
SCIENTIFIC REPORTS
(2021)
Article
Agronomy
Li-Wei Meng, Guo-Rui Yuan, Meng-Ling Chen, Wei Dou, Tian-Xing Jing, Li-Sha Zheng, Meng-Lan Peng, Wen-Jie Bai, Jin-Jun Wang
Summary: This study identified lncRNAs potentially related to malathion resistance in Bactrocera dorsalis, with two lncRNAs possibly influencing resistance by modulating the structure or components of the cuticle.
PEST MANAGEMENT SCIENCE
(2021)
Article
Biochemistry & Molecular Biology
Lanlan Wang, Siwen Wu, Jingjing Jin, Ran Li
Summary: Long non-coding RNAs (lncRNAs) are playing roles in plant defense against herbivores, as shown by the identification of 238 armyworm (AW)-elicited lncRNAs in rice. One of the differentially expressed lncRNAs was identified from antisense transcripts of the jasmonate ZIM-domain gene JAZ10.
PLANT SIGNALING & BEHAVIOR
(2021)
Article
Biochemical Research Methods
Sunyoung Kwon, Byunghan Lee, Sungroh Yoon
BMC BIOINFORMATICS
(2014)
Article
Biochemical Research Methods
Sunyoung Kwon, Sungroh Yoon
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2019)
Article
Biochemical Research Methods
Sunyoung Kwon, Ho Bae, Jeonghee Jo, Sungroh Yoon
BMC BIOINFORMATICS
(2019)
Article
Oncology
Han-Byoel Lee, Sae Byul Lee, Minsu Kim, Sunyoung Kwon, Jeonghee Jo, Jinkyoung Kim, Hee Jin Lee, Han-Suk Ryu, Jong Won Lee, Chungyeul Kim, Jaehwan Jeong, Hyoki Kim, Dong-Young Noh, In-Ae Park, Sei-Hyun Ahn, Sun Kim, Sungroh Yoon, Aeree Kim, Wonshik Han
CLINICAL CANCER RESEARCH
(2020)
Article
Computer Science, Artificial Intelligence
Kyung-Wha Park, Jung-Woo Ha, JungHoon Lee, Sunyoung Kwon, Kyung-Min Kim, Byoung-Tak Zhang
Summary: Evaluating advertisements based on user preferences and ad quality is crucial in marketing. Utilizing deep neural networks and auxiliary attributes can enhance ad image preferences and appeal. The proposed M2FN network achieves state-of-the-art performance in predicting user preferences using real-world ad datasets with rich auxiliary attributes.
APPLIED SOFT COMPUTING
(2021)
Article
Biochemical Research Methods
Kisung Moon, Hyeon-Jin Im, Sunyoung Kwon
Summary: Motivation Self-supervised learning (SSL) is a method that utilizes supervision inherent in the data to learn the data representation. In the drug field, SSL has shown excellent performance for molecular property prediction using unlabeled data. However, there are limitations with existing SSL models, such as their large-scale nature and lack of utilization of 3D structural information. We propose a novel contrastive learning framework, 3DGCL, to address these problems and achieve state-of-the-art performance in molecular property prediction.
Article
Computer Science, Information Systems
Dae-Il Noh, Seon-Geun Jeong, Huu-Trung Hoang, Quoc-Viet Pham, Thien Huynh-The, Mikio Hasegawa, Hiroo Sekiya, Sun-Young Kwon, Sang-Hwa Chung, Won-Joo Hwang
Summary: In this study, a radio frequency-based solution was proposed for detecting unauthorized drone use. By using a power-based spectrogram image with an applied threshold value for convolutional neural network training, the method demonstrated noise tolerance and scalability.
Article
Computer Science, Information Systems
Sunyoung Kwon, Gyuwan Kim, Byunghan Lee, Jongsik Chun, Sungroh Yoon, Young-Han Kim
Summary: Nucleic acid sequence classification is a fundamental task in bioinformatics, and NASUCP is a new method that captures the statistical structures of nucleotide sequences using compact context-tree models and universal probability from information theory. Experimental results show that NASUCP outperforms widely-used alternatives in efficiency, accuracy, and scalability, and can also be applied to other bioinformatics tasks.
Article
Computer Science, Information Systems
Kyuyong Shin, Wonyoung Shin, Jung-Woo Ha, Sunyoung Kwon
Summary: Graph neural networks have shown promising results in representing diverse graph-structured data, but face issues like oversmoothing and limited generalization. To address these challenges, we propose GESM model which considers relationships, features, and structure, achieving state-of-the-art or comparable performances on multiple tasks.
Article
Computer Science, Information Systems
Seongsil Heo, Sunyoung Kwon, Jaekoo Lee
Summary: This paper aims to improve stress-detection performance through precise signal processing based on PPG data, proposing a two-step denoising method and ensemble-based multiple peak-detecting method, achieving an accuracy of 96.50% and an F1 score of 93.36% on the WESAD dataset.
Article
Dermatology
Sunyoung Kwon, Ji Young Choi, Jung-Won Shin, Chang-Hun Huh, Kyoung-Chan Park, Mi-Hee Du, Sungroh Yoon, Jung-Im Na
ACTA DERMATO-VENEREOLOGICA
(2019)
Proceedings Paper
Mathematical & Computational Biology
Sunyoung Kwon, Sungroh Yoon
ACM-BCB' 2017: PROCEEDINGS OF THE 8TH ACM INTERNATIONAL CONFERENCE ON BIOINFORMATICS, COMPUTATIONAL BIOLOGY,AND HEALTH INFORMATICS
(2017)