THGB: predicting ligand-receptor interactions by combining tree boosting and histogram-based gradient boosting DOI Creative Commons
Liqian Zhou, Jiao Song, Zejun Li

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Nov. 28, 2024

Ligand-receptor interaction (LRI) prediction has great significance in biological and medical research facilitates to infer analyze cell-to-cell communication. However, wet experiments for new LRI discovery are costly time-consuming. Here, we propose a computational model called THGB uncover LRIs. first extracts feature information of Ligand-Receptor (LR) pairs using iFeature. Next, it adopts tree boosting obtain representative LR features. Finally, devises the histogram-based gradient capture high-quality To assess performance, compared with three models (i.e., CellEnBoost, CellGiQ, CellComNet) one classical protein-protein inference PIPR. The results demonstrated that achieved best overall predictions terms six evaluation indictors precision, recall, accuracy, F1-score, AUC, AUPR). measure effect selection on prediction, was four methods PCA, NMF, LLE, TSVD). showed more appropriate select features improve prediction. We also conducted ablation study found outperformed without selection. hope is useful tool find LRIs further

Language: Английский

Regulation-aware graph learning for drug repositioning over heterogeneous biological network DOI
Bo-Wei Zhao, Xiaorui Su,

Yue Yang

et al.

Information Sciences, Journal Year: 2024, Volume and Issue: 686, P. 121360 - 121360

Published: Aug. 22, 2024

Language: Английский

Citations

18
Predicting the potential associations between circRNA and drug sensitivity using a multisource feature‐based approach DOI Creative Commons

Shuaidong Yin,

Peng Xu, Yefeng Jiang

et al.

Journal of Cellular and Molecular Medicine, Journal Year: 2024, Volume and Issue: 28(19)

Published: Sept. 30, 2024

Abstract The unique non‐coding RNA molecule known as circular (circRNA) is distinguished from conventional linear by having a longer half‐life, greater degree of conservation and inherent solidity. Extensive research has demonstrated the profound impact circRNA expression on cellular drug sensitivity therapeutic efficacy. There an immediate need for creation efficient computational techniques to anticipate potential correlations between sensitivity, classical biological approaches are time‐consuming costly. In this work, we introduce novel deep learning model called SNMGCDA, which aims forecast relationships sensitivity. SNMGCDA incorporates diverse range similarity networks, enabling derivation feature vectors circRNAs drugs using three distinct calculation methods. First, utilize sparse autoencoder extraction characteristics. Subsequently, application non‐negative matrix factorization (NMF) enables identification based their shared features. Additionally, multi‐head graph attention network employed capture characteristics circRNAs. After acquiring these separate components, combine them form unified inclusive vector each cluster drug. Finally, relevant labels inputted into multilayer perceptron (MLP) make predictions. outcomes experiment, obtained through 5‐fold cross‐validation (5‐fold CV) 10‐fold (10‐fold CV), demonstrate outperforms five other state‐of‐art methods in terms performance. majority case studies have predominantly confirmed newly discovered thereby emphasizing its reliability predicting drugs.

Language: Английский

Citations

12
Identifying potential ligand–receptor interactions based on gradient boosted neural network and interpretable boosting machine for intercellular communication analysis DOI
Lihong Peng, Pengfei Gao, Wei Xiong

et al.

Computers in Biology and Medicine, Journal Year: 2024, Volume and Issue: 171, P. 108110 - 108110

Published: Feb. 6, 2024

Language: Английский

Citations

11
LncRNA–miRNA interactions prediction based on meta‐path similarity and Gaussian kernel similarity DOI Creative Commons

Jingxuan Xie,

Peng Xu,

Lin Ye

et al.

Journal of Cellular and Molecular Medicine, Journal Year: 2024, Volume and Issue: 28(19)

Published: Sept. 30, 2024

Abstract Long non‐coding RNAs (lncRNAs) and microRNAs (miRNAs) are two typical types of that interact play important regulatory roles in many animal organisms. Exploring the unknown interactions between lncRNAs miRNAs contributes to a better understanding their functional involvement. Currently, studying heavily relies on laborious biological experiments. Therefore, it is necessary design computational method for predicting lncRNA–miRNA interactions. In this work, we propose called MPGK‐LMI, which utilizes graph attention network (GAT) predict animals. First, construct meta‐path similarity matrix based known interaction information. Then, use GAT aggregate constructed computed Gaussian kernel update feature with neighbourhood Finally, scoring module used prediction. By comparing three state‐of‐the‐art algorithms, MPGK‐LMI achieves best results terms performance, AUC value 0.9077, AUPR 0.9327, ACC 0.9080, F1‐score 0.9143 precision 0.8739. These validate effectiveness reliability MPGK‐LMI. Additionally, conduct detailed case studies demonstrate feasibility our approach practical applications. Through these empirical results, gain deeper insights into mechanisms interactions, providing significant breakthroughs advancements field research. summary, not only outperforms others performance but also establishes its practicality research through real‐case analysis, offering strong support guidance future

Language: Английский

Citations

11
GEnDDn: An lncRNA–Disease Association Identification Framework Based on Dual-Net Neural Architecture and Deep Neural Network DOI
Lihong Peng,

Mengnan Ren,

Liangliang Huang

et al.

Interdisciplinary Sciences Computational Life Sciences, Journal Year: 2024, Volume and Issue: 16(2), P. 418 - 438

Published: May 11, 2024

Language: Английский

Citations

7
Finding potential lncRNA–disease associations using a boosting-based ensemble learning model DOI Creative Commons
Liqian Zhou,

Xinhuai Peng,

Lijun Zeng

et al.

Frontiers in Genetics, Journal Year: 2024, Volume and Issue: 15

Published: March 1, 2024

Introduction: Long non-coding RNAs (lncRNAs) have been in the clinical use as potential prognostic biomarkers of various types cancer. Identifying associations between lncRNAs and diseases helps capture design efficient therapeutic options for diseases. Wet experiments identifying these are costly laborious. Methods: We developed LDA-SABC, a novel boosting-based framework lncRNA–disease association (LDA) prediction. LDA-SABC extracts LDA features based on singular value decomposition (SVD) classifies pairs (LDPs) by incorporating LightGBM AdaBoost into convolutional neural network. Results: The performance was evaluated under five-fold cross validations (CVs) lncRNAs, diseases, LDPs. It obviously outperformed four other classical inference methods (SDLDA, LDNFSGB, LDASR, IPCAF) through precision, recall, accuracy, F1 score, AUC, AUPR. Based accurate prediction we used it to find lncRNA lung results elucidated that 7SK HULC could relationship with non-small-cell cancer (NSCLC) adenocarcinoma (LUAD), respectively. Conclusion: hope our proposed method can help improve identification.

Language: Английский

Citations

5
GraphADT: empowering interpretable predictions of acute dermal toxicity with multi-view graph pooling and structure remapping DOI Creative Commons
Xinqian Ma, Xiangzheng Fu, Tao Wang

et al.

Bioinformatics, Journal Year: 2024, Volume and Issue: 40(7)

Published: July 1, 2024

Abstract Motivation Accurate prediction of acute dermal toxicity (ADT) is essential for the safe and effective development contact drugs. Currently, graph neural networks, a form deep learning technology, accurately model structure compound molecules, enhancing predictions their ADT. However, many existing methods emphasize atom-level information transfer overlook crucial data conveyed by molecular bonds interrelationships. Additionally, these often generate “equal” node representations across entire graph, failing to accentuate “important” substructures like functional groups, pharmacophores, toxicophores, thereby reducing interpretability. Results We introduce novel model, GraphADT, utilizing remapping multi-view pooling (MVPool) technologies predict Initially, our applies better delineate bonds, transforming “bonds” into new nodes “bond-atom-bond” interactions edges, reconstructing graph. Subsequently, we use MVPool amalgamate from various perspectives, minimizing biases inherent single-view analyses. Following this, generates robust ranking collaboratively, emphasizing critical or enhance Lastly, apply comparison strategy train both original remapped graphs, deriving final representation. Experimental results on public datasets indicate that GraphADT outperforms state-of-the-art models. The has been demonstrated effectively ADT, offering potential guidance drugs related treatments. Availability implementation Our code are accessible at: https://github.com/mxqmxqmxq/GraphADT.git.

Language: Английский

Citations

5
Accurate identification of snoRNA targets using variational graph autoencoder to advance the redevelopment of traditional medicines DOI Creative Commons
Zhina Wang,

Yishan Chen,

Hongming Ma

et al.

Frontiers in Pharmacology, Journal Year: 2025, Volume and Issue: 15

Published: Jan. 6, 2025

Existing studies indicate that dysregulation or abnormal expression of small nucleolar RNA (snoRNA) is closely associated with various diseases, including lung cancer. Furthermore, these diseases often involve multiple targets, making the redevelopment traditional medicines highly promising. Accurate prediction potential snoRNA therapeutic targets essential for early disease intervention and medicines. Additionally, researchers have developed artificial intelligence (AI)-based methods to screen predict thereby advancing drug redevelopment. However, existing face challenges such as imbalanced datasets dominance high-degree nodes in graph neural networks (GNNs), which compromise accuracy node representations. To address challenges, we propose an AI model based on variational autoencoders (VGAEs) integrates decoupling Kolmogorov-Arnold Network (KAN) technologies. The reconstructs snoRNA-disease graphs by learning representations, accurately identifying targets. By similarity from degree, mitigates nodes, enhances scenarios like cancer, leverages KAN technology improve adaptability flexibility new data. Case revealed SNORA21 SNORD33 are abnormally expressed cancer patients strong candidates These findings validate proposed model's effectiveness supporting screening treatment, Data experimental archived in: https://github.com/shmildsj/data.

Language: Английский

Citations

0
MPEMDA: A Multi-Similarity Integration Approach with Pre-completion and Error Correction for Predicting Microbe-Drug Associations DOI
Yuxiang Li, Haochen Zhao, Jianxin Wang

et al.

Methods, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

Language: Английский

Citations

0
Predicting cell–cell communication by combining heterogeneous ensemble deep learning and weighted geometric mean DOI
Lihong Peng, L. C. Liu,

Liangliang Huang

et al.

Applied Soft Computing, Journal Year: 2025, Volume and Issue: unknown, P. 112839 - 112839

Published: Feb. 1, 2025

Language: Английский

Citations

0