Article
Biochemical Research Methods
Yulian Ding, Li-Ping Tian, Xiujuan Lei, Bo Liao, Fang-Xiang Wu
Summary: The critical roles of miRNAs in biological processes and diseases have been demonstrated, and exploring their relationships can aid in understanding disease mechanisms. This study introduces a VGAE-MDA framework for miRNA-disease association prediction, outperforming existing methods.
Article
Biochemical Research Methods
Zhengwei Li, Jiashu Li, Ru Nie, Zhu-Hong You, Wenzheng Bao
Summary: The abnormal expression of miRNAs is associated with the evolution and progression of various human complex diseases. Identifying disease-related miRNAs as new biomarkers can promote the development of disease pathology and clinical medicine. Designing biological experiments to validate disease-related miRNAs is usually time-consuming and expensive, highlighting the need for effective computational methods to predict potential miRNA-disease associations.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Xuan Huang, Zhenlong Hu, Lin Lin
Summary: Deep clustering combines deep learning and clustering to achieve superior clustering performance. This paper proposes an embedded auto-encoder network model that effectively encodes input object features and improves clustering through smoothness constraints.
Article
Computer Science, Artificial Intelligence
Jinyu Cai, Shiping Wang, Wenzhong Guo
Summary: The paper proposes a deep stacked sparse embedded clustering method that considers both local structure preservation and input sparsity. The deep learning approach jointly learns clustering-oriented features and optimizes cluster label assignments by minimizing both the reconstruction and clustering loss. Comprehensive experiments validate the effectiveness of introducing sparsity and preserving local structure in the proposed method.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Biochemical Research Methods
Huizhe Zhang, Juntao Fang, Yuping Sun, Guobo Xie, Zhiyi Lin, Guosheng Gu
Summary: This study proposes a novel method called AGAEMD to predict potential miRNA disease associations. It utilizes a node-level attention graph auto-encoder to represent nodes and calculate association scores. Experimental results demonstrate the excellent performance of AGAEMD compared to other methods, and case studies confirm its reliable predictive performance.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Computer Science, Information Systems
Yi Zhu, Lei Li, Xindong Wu
Summary: The research proposes a semi-supervised deep learning framework to address the issue of insufficient labeled image data, utilizing stacked layers, convolutional approach, and sparse auto-encoder to learn feature representations. The framework also includes an algorithm to handle data redundancy and encodes label information using a Softmax regression model.
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
(2021)
Article
Biochemical Research Methods
Qi Liang, Wenxiang Zhang, Hao Wu, Bin Liu
Summary: Discovering the relationships between long non-coding RNAs (lncRNAs) and diseases is significant in the treatment, diagnosis and prevention of diseases. Therefore, it is important to develop an efficient computational method for predicting potential lncRNA-disease associations. The proposed GraLTR-LDA predictor, based on biological knowledge graphs and ranking framework, outperforms other predictors and effectively detects potential lncRNA-disease associations.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Hu Lu, Saixiong Liu, Hui Wei, Chao Chen, Xia Geng
Summary: The study introduces a deep learning clustering model called DMACN for clustering brain disease functional connectivity data. Experimental results show that this algorithm performs well in clustering brain functional connectivity data compared to existing algorithms.
Article
Computer Science, Information Systems
Ruina Bai, Ruizhang Huang, Yongbin Qin, Yanping Chen, Chuan Lin
Summary: With the development of the internet, multi-view text documents have become common and research on multi-view document modeling has increased. Traditional single-view document modeling treats each document independently, while multi-view text documents have complex correlations. This study introduces a deep generative model called Hierarchical Variational Auto-Encoder (HVAE) that combines the advantages of probability generative models and deep neural networks. The proposed method successfully captures both global and local topical information in multi-view documents.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Yunhe Zhang, Zhoumin Lu, Shiping Wang
Summary: Feature selection is crucial in machine learning, and the proposed unsupervised feature selection scheme based on auto-encoder can solve traditional constrained feature selection problems and adapt to various loss functions and activation functions.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Wei Song, Yuxuan Zhang, Soon Cheol Park
Summary: In this paper, an energy and label constrained deep auto-encoder (ELDAE) is proposed to improve feature extraction ability for classification. Experimental results demonstrate that ELDAE outperforms six state-of-the-art algorithms in terms of classification accuracy.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Biochemical Research Methods
Ngan Dong, Stefanie Muecke, Megha Khosla
Summary: Growing evidence suggests that microRNAs could serve as biomarkers for complex human diseases. Machine learning techniques for miRNA-disease association prediction have gained attention due to the expensive and time-consuming nature of wet-lab experiments. However, data scarcity poses a challenge in building reliable machine learning models. The proposed MuCoMiD approach overcomes the limitations of existing models by incorporating knowledge from multiple biological information sources in a multitask setting.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Biology
Yao Fu, Runtao Yang, Lina Zhang
Summary: This paper proposes a new predictor, VGAERF, which combines VGAE and RF to explore the molecular mechanism of circRNA-disease associations and improve prediction performance, contributing to the diagnosis of circRNA-related diseases.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Biochemical Research Methods
Wei Liu, Hui Lin, Li Huang, Li Peng, Ting Tang, Qi Zhao, Li Yang
Summary: In this study, a new computational method called DFELMDA is proposed to predict miRNA-disease associations using deep forest ensemble learning and autoencoder. Results from experiments on the HMDD dataset show that DFELMDA outperforms other methods in terms of performance.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Computer Science, Information Systems
Wei Song, Wei Li, Ziyu Hua, Fuxin Zhu
Summary: The paper introduces a new Variable Learning Speed Deep Auto-Encoder (VLSDAE) that adaptsively adjusts its learning rate using Multiscale Reconstruction Errors (MRE) and Weights Update Correlation (WUC). The complexity analysis and theoretical proof demonstrate the convergence of VLSDAE, which outperforms state-of-the-art algorithms in terms of training error, classification accuracy, precision, recall and macro-F1 in experimental results.
INFORMATION SCIENCES
(2021)
Article
Biochemical Research Methods
Han-Yuan Zhang, Lei Wang, Zhu-Hong You, Lun Hu, Bo-Wei Zhao, Zheng-Wei Li, Yang-Ming Li
Summary: Researchers have discovered a novel topology of RNA transcript called circular RNA (circRNA) that competes with messenger RNA (mRNA) and long noncoding RNA in gene regulation. This finding suggests that circRNA could be associated with complex diseases, thus identifying the relationship between them would contribute to medical research. However, in vitro experiments to determine the circRNA-disease association are time-consuming and lack direction. To address this, a computational method called iGRLCDA was proposed, which utilizes graph convolution network (GCN) and graph factorization (GF) to predict circRNA-disease associations. The performance of iGRLCDA was compared to other prediction models using five-fold cross-validation, showing strong competitiveness and high accuracy.
BRIEFINGS IN BIOINFORMATICS
(2022)
Review
Computer Science, Information Systems
Zhanheng Chen, Zhuhong You, Qinhu Zhang, Zhenhao Guo, Siguo Wang, Yanbin Wang
Summary: This review provides a comprehensive overview of recent literature on computational prediction of self-interacting proteins (SIPs), serving as a valuable reference for future work. The review first describes the data required for predicting drug-target interactions (DTIs), followed by the presentation of interesting feature extraction methods and computational models. An empirical comparison is then conducted to demonstrate the prediction performance of various classifiers under different feature extraction and encoding schemes. Overall, potential methods for further enhancing SIPs prediction performance and related research directions are summarized and highlighted.
FRONTIERS OF COMPUTER SCIENCE
(2023)
Article
Biochemical Research Methods
Jing Li, Chen Zhang, Zhengwei Li, Ru Nie, Pengyong Han, Wenjia Yang, Hongmei Liao
Summary: This study proposes a novel model called GCMCDTI for predicting drug-target interactions using graph convolutional network based on matrix completion, achieving high AUC values on four benchmark datasets.
JOURNAL OF BIOINFORMATICS AND COMPUTATIONAL BIOLOGY
(2022)
Article
Biochemical Research Methods
Kangwei Wang, Zhengwei Li, Zhu-Hong You, Pengyong Han, Ru Nie
Summary: In this study, we propose an adversarial dense graph convolutional network architecture for feature learning in single-cell classification. We introduce dense connectivity mechanism and attention-based feature aggregation to enhance the representation of higher-order features and the organic combination between features. A feature reconstruction module is used to preserve the features of the original data and assist in single-cell classification. Experimental results show that our model outperforms existing classical methods in terms of classification accuracy on benchmark datasets.
Article
Biochemical Research Methods
Li Zhang, Chun-Chun Wang, Xing Chen
Summary: This study presents a novel model called MRBDTA to improve the existing computational models for drug-target binding affinity prediction. MRBDTA achieves better performance in prediction accuracy and can provide interpretability analysis. The case studies demonstrate the reliable performance of MRBDTA in drug design for SARS-CoV-2.
BRIEFINGS IN BIOINFORMATICS
(2022)
Editorial Material
Biochemical Research Methods
Xing Chen, Li Huang
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Automation & Control Systems
Lei Wang, Zhu-Hong You, De-Shuang Huang, Jian-Qiang Li
Summary: This study presents a new computational model, MGRCDA, which utilizes metagraph recommendation theory to predict potential circRNA-disease associations. By integrating heterogeneous biological networks and utilizing an iterative search algorithm, MGRCDA achieved high prediction accuracy and reliability. The experimental results demonstrate its feasibility and efficiency in reducing the scope of wet-lab experiments.
IEEE TRANSACTIONS ON CYBERNETICS
(2023)
Article
Biology
Yong-Jian Guan, Chang-Qing Yu, Yan Qiao, Li-Ping Li, Zhu-Hong You, Zhong-Hao Ren, Yue-Chao Li, Jie Pan
Summary: This study presents a computational method called MFIDMA for predicting drug-miRNA associations. The proposed model demonstrates excellent performance in experiments and can be used for the development and research of miRNA-targeted drugs, providing new perspectives on miRNA therapeutics research and drug discovery.
Article
Biochemical Research Methods
Chen-Di Han, Chun-Chun Wang, Li Huang, Xing Chen
Summary: Adverse drug-drug interactions (DDIs) have become a serious problem in healthcare. Researchers have proposed a Multi-Channel Feature Fusion model for multi-typed DDI prediction (MCFF-MTDDI), which effectively fuses different features extracted from drug chemical structure, drug pairs' extra label, and drug knowledge graph (KG) to predict multi-typed DDIs. The results of experiments on multiple datasets demonstrate the effectiveness of MCFF-MTDDI.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biotechnology & Applied Microbiology
Hua Hu, Huan Zhao, Tangbo Zhong, Xishang Dong, Lei Wang, Pengyong Han, Zhengwei Li
Summary: In this study, a novel miRNA-disease association prediction model (ADPMDA) based on adaptive deep propagation graph neural network is proposed. The miRNA-disease heterogeneous graph is constructed, and the features of miRNAs and diseases are projected into a low-dimensional space. Attention mechanism is used to aggregate local features of central nodes. The effectiveness of ADPMDA is demonstrated through experiments on human miRNA disease database v3.0 dataset.
BRIEFINGS IN FUNCTIONAL GENOMICS
(2023)
Article
Biochemical Research Methods
Huan Zhao, Zhengwei Li, Zhu-Hong You, Ru Nie, Tangbo Zhong
Summary: Numerous experiments have demonstrated the abnormal expression of microRNA (miRNA) in complex human diseases. Identifying the associations between miRNAs and diseases is crucial for clinical medicine, but traditional experimental methods are often inefficient. Therefore, a deep learning method called NSAMDA, based on neighbor selection graph attention networks, is proposed to predict miRNA-disease associations. The NSAMDA model achieves satisfactory performance in predicting miRNA-disease associations, surpassing the most advanced model, as demonstrated through experiments on various diseases.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Jie Pan, Zhuhong You, Wencai You, Tian Zhao, Chenlu Feng, Xuexia Zhang, Fengzhi Ren, Sanxing Ma, Fan Wu, Shiwei Wang, Yanmei Sun
Summary: This study developed a model called PTBGRP based on microbial heterogeneous interaction network to predict new phages for bacterial hosts. By integrating different biological attributes and topological features, a deep neural network classifier was used to predict unknown PBI pairs. Experimental results demonstrated that PTBGRP achieved the best performance on pathogen and PBI datasets.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Biology
Lihong Peng, Jingwei Tan, Wei Xiong, Li Zhang, Zhao Wang, Ruya Yuan, Zejun Li, Xing Chen
Summary: The study introduces a new deep learning framework, CellComNet, which deciphers cell-cell communication mediated by extracellular molecules through the analysis of single-cell transcriptomic data. The framework demonstrates efficient identification of credible LRIs and significantly improves the inference performance of cell-cell communication. It has the potential to contribute to anticancer drug design and tumor-targeted therapy.
COMPUTERS IN BIOLOGY AND MEDICINE
(2023)
Article
Biochemical Research Methods
Lihong Peng, Ruya Yuan, Chendi Han, Guosheng Han, Jingwei Tan, Zhao Wang, Min Chen, Xing Chen
Summary: Cell-to-cell communication (CCC) plays significant roles in multicellular organisms, especially in cancer genesis, development, and metastasis. This manuscript presents a Boosting-based LRI identification model (CellEnBoost) for predicting and interpreting ligand-receptor interactions in CCC. Experimental results demonstrate the superior performance of this model and its validation in human head and neck squamous cell carcinoma tissues.
IEEE TRANSACTIONS ON NANOBIOSCIENCE
(2023)
Article
Biochemical Research Methods
Zheng-Wei Li, Qian-Kun Wang, Chang-An Yuan, Peng-Yong Han, Zhu-Hong You, Lei Wang
Summary: Growing studies have shown the close association between miRNAs and human diseases, and computational approaches have achieved promising results in predicting miRNA-disease associations. In this article, a graph representation learning method is proposed for predicting miRNA-disease associations, and the effectiveness of the model is validated through five-fold cross-validation.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)