Article
Biochemical Research Methods
Jing-Zhe Jiang, Wen-Guang Yuan, Jiayu Shang, Ying-Hui Shi, Li-Ling Yang, Min Liu, Peng Zhu, Tao Jin, Yanni Sun, Li-Hong Yuan
Summary: The paper presents PhaGCN2, a tool that rapidly classifies viral sequences at the family level and improves the precision and recall of virus classification. It allows for high-throughput processing of viral sequences and supports visualization.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biology
Guangqi Wen, Peng Cao, Huiwen Bao, Wenju Yang, Tong Zheng, Osmar Zaiane
Summary: In this study, a machine learning approach is proposed for the classification of neurological disorders, with an interpretable framework. The results of experiments demonstrate the high classification performance of the proposed method on datasets of Autism Spectrum Disorders and Alzheimer's disease.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Feng Zhao, Junjie Zhang, Zhe Meng, Hanqiang Liu, Zhenhui Chang, Jiulun Fan
Summary: This article proposes a hybrid network called MVAHN for hyperspectral image (HSI) classification, which combines convolutional neural network (CNN) and transformer structures. It also utilizes a graph convolutional module (GCM) to extract multiple types of feature information. Experimental results show that MVAHN achieves high accuracy on various datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Mathematics
Mohammad Abrar Shakil Sejan, Md Habibur Rahman, Md Abdul Aziz, Jung-In Baik, Young-Hwan You, Hyoung-Kyu Song
Summary: In this paper, a graph convolutional network (GCN) model is proposed to enhance the performance of node classification tasks. A GCN layer is designed by updating the aggregation function using an updated value of the weight coefficient. The aggregation function is calculated using the adjacency matrix of the input graph and the identity matrix. Extensive experimental studies with seven publicly available datasets validate the proposed model, which achieves comparable results with the state-of-the-art methods. The proposed approach can achieve superior results with one single layer.
Article
Computer Science, Interdisciplinary Applications
Xuefen Lin, Jielin Chen, Weifeng Ma, Wei Tang, Yuchen Wang
Summary: This study proposes an improved graph convolution model that achieves effective emotion classification in complex dataset environments and reduces the cost of affective computing.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Engineering, Biomedical
Xiaoyi Chen, Jing Zhou, Pengfei Ke, Jiayuan Huang, Dongsheng Xiong, Yuanyuan Huang, Guolin Ma, Yuping Ning, Fengchun Wu, Kai Wu
Summary: Recent studies have shown that brain connectivity abnormalities are associated with schizophrenia. However, most previous studies using machine learning methods focused on MRI features of brain regions, ignoring brain connectivity and its network topology. In this study, we used a graph convolutional network (GCN) to classify schizophrenia patients based on both brain region and connectivity features derived from combined functional MRI and connectomics analysis. The results showed that the proposed method significantly improved classification performance compared to traditional machine learning and deep learning methods based on MRI features alone.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Computer Science, Artificial Intelligence
Ahmed A. Mubarak, Han Cao, Ibrahim M. Hezam, Fei Hao
Summary: This study proposes a model based on Graph Convolutional Network to classify students into different behavioral patterns based on their engagement with course videos. By introducing a label function to label datasets, the model provides a learning foundation for future data processing. Experimental results show that the proposed model achieves high predictive accuracy and f1-score.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Youyong Kong, Jiaxing Li, Ke Zhang, Jiasong Wu
Summary: Data augmentation can enhance the generalization performance of neural networks, but it is challenging for graph data due to its irregular non-Euclidean structure. This paper introduces MSSA-Mixup, a novel graph data augmentation method that extends the training distribution through interpolating multi-scale graph representation with self-attention. MSSA-Mixup effectively improves the generalization ability of GNNs, as demonstrated by extensive experiments on benchmark datasets.
PATTERN RECOGNITION LETTERS
(2023)
Article
Multidisciplinary Sciences
Yong Chen, Xiao-Zhu Xie, Wei Weng, Yi-Fan He
Summary: In graph-structured data, effectively utilizing the rich information in node content is crucial to improve the performance of graph convolutional networks (GCNs). We propose a novel graph attention network with adaptability that fully utilizes multi-order content and combines high- and low-order attention mechanisms. Experimental results on multiple datasets validate the feasibility and effectiveness of the proposed model.
Article
Automation & Control Systems
Hao Wang, Wenchuan Yang, Jichao Li, Junwei Ou, Yanjie Song, Yingwu Chen
Summary: This paper proposes an improved semi-supervised heterogeneous graph convolutional network model, IHGCN, for job recommendation. The model constructs a heterogeneous resume graph and learns node representation to provide effective job recommendations. Experimental results show that IHGCN outperforms the baselines.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Albert Mosella-Montoro, Javier Ruiz-Hidalgo
Summary: This paper presents a 2D-3D Fusion stage that combines 3D Geometric Features with 2D Texture Features to achieve a more robust geometric embedding, outperforming the current state-of-the-art in RGB-D indoor scene classification task based on experimental results using NYU-Depth-V2 and SUN RGB-D datasets.
INFORMATION FUSION
(2021)
Article
Geochemistry & Geophysics
Bing Yang, Hailiang Ye, Ming Li, Feilong Cao, Shirui Pan
Summary: This article proposes a Global-to-Local decoupling Graph network (GoLoG) for hyperspectral image (HSI) classification. It optimizes both the graph structure and network parameters to capture long-range spatial correlations in HSI while preserving individualized spectral characteristics of each pixel. The proposed GoLoG outperforms other state-of-the-art HSI classification methods.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Remote Sensing
Huafei Yu, Tinghua Ai, Min Yang, Lina Huang, Jiaming Yuan
Summary: Drainage pattern recognition (DPR) is a challenging problem in various fields. In this study, a graph convolutional network (GCN) was introduced to improve traditional DPR methods by considering manual recognition effects. The proposed approach achieved higher accuracy compared to other machine learning methods, demonstrating the potential and room for improvement of GCN in DPR.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2022)
Article
Mathematics
Seonghun Kim, Seockhun Bae, Yinhua Piao, Kyuri Jo
Summary: Integrating gene expression data and biological networks into the analysis framework for drug response prediction can improve prediction accuracy. DrugGCN successfully achieves this goal through graph convolutional network technology and demonstrates its success in biological data.
Article
Computer Science, Artificial Intelligence
Sheng Wan, Shirui Pan, Shengwei Zhong, Jie Yang, Jian Yang, Yibing Zhan, Chen Gong
Summary: In this paper, a Multi-level Graph Learning Network (MGLN) is proposed for hyperspectral image (HSI) classification. MGLN can learn the local and global structural information of the graph in an end-to-end manner, and encode spatial relevance and global contextual information through attention mechanism and feature representations. Experimental results demonstrate that MGLN outperforms existing methods on real-world datasets.
PATTERN RECOGNITION
(2022)
Article
Environmental Sciences
Yong Wang, Yanni Sun, Le Chen, Hua Shao, Yanhua Zeng, Yongjun Zeng, Feiyu Tang, Junhuo Cai, Shan Huang
Summary: Rice agriculture is a significant source of methane emissions and cadmium accumulation. The study investigated the combined effects of water management and lime application on CH4 emissions and rice Cd uptake. Results showed that flooding following midseason drainage effectively reduced CH4 emissions, while lime application reduced both CH4 emissions and rice Cd uptake. The recommended approach to mitigate CH4 emissions without increasing Cd uptake is continuous flooding with midseason drainage combined with lime application.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Biochemical Research Methods
Jing-Zhe Jiang, Wen-Guang Yuan, Jiayu Shang, Ying-Hui Shi, Li-Ling Yang, Min Liu, Peng Zhu, Tao Jin, Yanni Sun, Li-Hong Yuan
Summary: The paper presents PhaGCN2, a tool that rapidly classifies viral sequences at the family level and improves the precision and recall of virus classification. It allows for high-throughput processing of viral sequences and supports visualization.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Virology
Fan Zhang, Andrew Gia, Guowei Chen, Lan Gong, Jason Behary, Georgina L. Hold, Amany Zekry, Xubo Tang, Yanni Sun, Emad El-Omar, Xiao-Tao Jiang
Summary: Viruses are abundant and important in ecosystems, and the two methods of whole genome shotgun metagenome and viral-like particle enriched metagenome sequencing are commonly used for comparing viruses in different environments. In this study, both methods were applied to investigate the stool virome in HCC patients and healthy controls, and they both successfully identified altered viral profiles in HCC patients. Ultra-deep sequencing recovered more viruses, and VLPM method can detect RNA viruses. Using both methods would capture different parts of the total virome and identify shared and specific viral signatures.
Article
Multidisciplinary Sciences
Zhixin Ma, Yifeng Liu, Lin Chai, Guochen Jin, Yanni Sun, Shaomin Zhou, Peiyuan Yin, Siwen Wang, Yuning Zhu, Dan Zhang, Shiming Lu, Bo Zhu
Summary: Intrahepatic cholestasis of pregnancy (ICP) is a rare liver disease characterized by disordered bile acid metabolism during pregnancy. Different types of ICP show distinct bile acid metabolism profiles. Elevated levels of total bile acids and glycocholic acid are associated with preterm birth in early-onset ICP (EICP), while increased levels of total bile acids and taurocholic acid are predictive of preterm birth in late-onset ICP (LICP). This study highlights the importance of assessing bile acid metabolism in ICP patients to predict perinatal complications.
SCIENTIFIC REPORTS
(2023)
Article
Biochemical Research Methods
Yongxin Ji, Jiayu Shang, Xubo Tang, Yanni Sun
Summary: Understanding the host range of plasmids is essential for studying their roles in bacterial evolution and adaptation. Existing tools for predicting plasmid hosts face challenges in sensitivity and precision. This study presents a hierarchical classification tool called HOTSPOT, which uses a state-of-the-art language model, Transformer, to accurately predict the host taxonomy of input plasmid contigs.
Article
Chemistry, Multidisciplinary
Qing Feng, Xianchao Chen, Ning Zhang, Xiaonan Li, Jingchao Zhou, Shengsheng Li, Xiaorong Zhang, Yanni Sun, Yuehui She
Summary: Bohai Oilfield has developed a bio-nano-depressurization and injection-increasing composite system solution to address the high injection pressure and insufficient injection volume in offshore oilfields. The new technology has the advantages of efficient decompression, long-term injection, and wide adaptation. However, there is a need for optimization schemes and application effect prediction methods to further promote and apply the bio-nano-composite system solution. This paper optimizes the injection volume, concentration, and speed of the bio-nano-augmentation fluid and evaluates the application effect using well testing, water absorption index, and numerical simulation methods.
Article
Biochemical Research Methods
Jiayu Shang, Cheng Peng, Xubo Tang, Yanni Sun
Summary: In this study, a computational method called PhaVIP was developed for fast and accurate classification and annotation of phage virion proteins (PVPs). By encoding protein sequences into unique images and utilizing the Vision Transformer model, PhaVIP can learn both local and global features from sequence images. Experimental results demonstrated the superior performance of PhaVIP, and its output was further applied to phage taxonomy classification and phage host prediction with beneficial results.
Article
Biochemistry & Molecular Biology
Xubo Tang, Jiayu Shang, Yongxin Ji, Yanni Sun
Summary: In this study, a plasmid detection tool called PLASMe was developed, which combines alignment and learning-based methods to effectively identify closely related and diverged plasmids. By encoding plasmid sequences as a language defined on the protein cluster-based token set, the Transformer model in PLASMe can learn the importance of proteins and their correlation. Comparative analysis showed that PLASMe achieved the highest F1-score in detecting complete plasmids, plasmid contigs, and contigs assembled from CAMI2 simulated data, and exhibited more reliable performance than other tools.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Microbiology
Herui Liao, Yongxin Ji, Yanni Sun
Summary: In this study, a new strain-level composition analysis tool named StrainScan is introduced, which employs a novel tree-based k-mers indexing structure to strike a balance between strain identification accuracy and computational complexity. Extensive testing on simulated and real sequencing data shows that StrainScan outperforms popular strain-level analysis tools in terms of accuracy and resolution. It provides more informative strain composition analysis in one sample or across multiple samples.
Article
Multidisciplinary Sciences
Weidun Xie, Xingjian Chen, Zetian Zheng, Fuzhou Wang, Xiaowei Zhu, Qiuzhen Lin, Yanni Sun, Ka-Chun Wong
Summary: This study presents a method called lncRNA-Top to predict lncRNA-gene regulation relationships and constructs controlled deep-learning models. Through case studies, it is found that the predictions are accurate, and additional software is provided for target candidate annotation.
Article
Biochemical Research Methods
Jiayu Shang, Xubo Tang, Yanni Sun
Summary: Researchers have developed a tool called PhaTYP that accurately predicts the lifestyle of bacteriophages, especially for short contigs. Experimental results show that PhaTYP outperforms other existing methods and achieves more stable performance on short contigs. Additionally, the utility of PhaTYP for analyzing phage lifestyle in human neonates' gut data has been demonstrated, which helps extend our understanding of microbial communities.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Runzhou Yu, Dehan Cai, Yanni Sun
Summary: RNA viruses mutate constantly, but accurate assembly of viral genomes is crucial for studying virus evolution and understanding the relationship between genotypes and virus properties. A new tool called AccuVIR has been developed for viral genome assembly and polishing using error-prone long reads. It can distinguish sequencing errors from true variants, resulting in more accurate viral genomes compared to other tools.
Article
Biochemical Research Methods
Guowei Chen, Xubo Tang, Mang Shi, Yanni Sun
Summary: In this study, we developed VirBot, a simple yet effective RNA virus identification tool based on protein families and adaptive score cutoffs. Compared to seven popular tools for virus identification, VirBot demonstrated high specificity in metagenomic datasets and superior sensitivity in detecting novel RNA viruses on both simulated and real sequencing data.
Article
Engineering, Environmental
Donglin Wang, Jiayu Shang, Hui Lin, Jinsong Liang, Chenchen Wang, Yanni Sun, Yaohui Bai, Jiuhui Qu
Summary: This study develops a bio-informatic pipeline using deep learning techniques to identify phages carrying antibiotic resistance genes (ARGs) and predict their hosts, with a focus on pathogens. The study discovers that temperate phages in a landscape lake replenished by reclaimed water predominantly carry ARGs related to multidrug resistance and beta-lactam antibiotics. In silico analysis and qPCR confirm a positive correlation between temperate phages and host pathogens, and seasonal variations in the abundance of phages and chromosomes carrying ARGs.
Article
Soil Science
Ping Liao, Lei Liu, Jin Chen, Yanni Sun, Shan Huang, Yongjun Zeng, Kees Jan van Groenigen
Summary: Liming materials can increase rice yield and nitrogen uptake, but decrease the efficiency of fertilizer nitrogen and promote nitrogen losses. Long-term studies on the impact of liming on nitrogen dynamics in paddy soils are necessary.
SOIL & TILLAGE RESEARCH
(2024)