| 1. | Massimo Orazio Spata; Sabrina Conoci; Alessandro Ortis; Sebastiano Battiato: A Novel Pipeline Based on Attention Mechanism Within a GNN to Predict New Edges Between Diseases and Genes. 2024, ISBN: 979-8-3503-7800-9. (Type: Proceeding | Abstract | Links | BibTeX) @proceedings{Battiato2024, title = {A Novel Pipeline Based on Attention Mechanism Within a GNN to Predict New Edges Between Diseases and Genes}, author = {Massimo Orazio Spata; Sabrina Conoci; Alessandro Ortis; Sebastiano Battiato}, editor = {IEEE}, url = {https://ieeexplore.ieee.org/abstract/document/10796979}, doi = {10.1109/MetroXRAINE62247.2024.10796979}, isbn = {979-8-3503-7800-9}, year = {2024}, date = {2024-12-24}, booktitle = {2024 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE)}, pages = {565-569}, abstract = {In recent years, Graph Neural Networks (GNNs) have shown great success in predicting missing edges in complex networks such as social networks, protein-protein interaction networks, and knowledge graphs, among others. One of the areas where GNNs may be applied is in predicting new edges between diseases and genes. The relationship between diseases and genes is complex, with a single gene being associated with multiple diseases, and multiple genes being associated with a single disease. GNNs are well-suited to capture such complex relationships as they can incorporate the structural information of the graph along with node and edge features to make predictions. Nodes represent diseases and genes, whereas edges represent known associations between them. Node features can be defined based on various characteristics of diseases and genes, such as Gene Ontology (GO) terms, pathways, and gene expression profiles. Edge features can be defined based on the strength of the association be-tween diseases and genes. The presented research shows a novel pipeline which integrating attention mechanisms into GNNs enhancing model's ability to focus on relevant gene-disease associations, and improving prediction accuracy. Experimental results exhibit a peak accuracy of 80%, highlighting the effectiveness and precision of the approach. This notable accuracy emphasizes the reliability and real-world relevance of the findings for predicting new edges between diseases and genes.}, keywords = {}, pubstate = {published}, tppubtype = {proceedings} } In recent years, Graph Neural Networks (GNNs) have shown great success in predicting missing edges in complex networks such as social networks, protein-protein interaction networks, and knowledge graphs, among others. One of the areas where GNNs may be applied is in predicting new edges between diseases and genes. The relationship between diseases and genes is complex, with a single gene being associated with multiple diseases, and multiple genes being associated with a single disease. GNNs are well-suited to capture such complex relationships as they can incorporate the structural information of the graph along with node and edge features to make predictions. Nodes represent diseases and genes, whereas edges represent known associations between them. Node features can be defined based on various characteristics of diseases and genes, such as Gene Ontology (GO) terms, pathways, and gene expression profiles. Edge features can be defined based on the strength of the association be-tween diseases and genes. The presented research shows a novel pipeline which integrating attention mechanisms into GNNs enhancing model's ability to focus on relevant gene-disease associations, and improving prediction accuracy. Experimental results exhibit a peak accuracy of 80%, highlighting the effectiveness and precision of the approach. This notable accuracy emphasizes the reliability and real-world relevance of the findings for predicting new edges between diseases and genes. |
2024 |
Massimo Orazio Spata; Sabrina Conoci; Alessandro Ortis; Sebastiano Battiato 2024, ISBN: 979-8-3503-7800-9. Abstract | Links | BibTeX | Tags: Accuracy, Attention mechanisms, Diseases, Graph neural networks, Ontologies, Pipelines, Predictive models, Proteins, Reliability, Social networking (online) @proceedings{Battiato2024, title = {A Novel Pipeline Based on Attention Mechanism Within a GNN to Predict New Edges Between Diseases and Genes}, author = {Massimo Orazio Spata; Sabrina Conoci; Alessandro Ortis; Sebastiano Battiato}, editor = {IEEE}, url = {https://ieeexplore.ieee.org/abstract/document/10796979}, doi = {10.1109/MetroXRAINE62247.2024.10796979}, isbn = {979-8-3503-7800-9}, year = {2024}, date = {2024-12-24}, booktitle = {2024 IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering (MetroXRAINE)}, pages = {565-569}, abstract = {In recent years, Graph Neural Networks (GNNs) have shown great success in predicting missing edges in complex networks such as social networks, protein-protein interaction networks, and knowledge graphs, among others. One of the areas where GNNs may be applied is in predicting new edges between diseases and genes. The relationship between diseases and genes is complex, with a single gene being associated with multiple diseases, and multiple genes being associated with a single disease. GNNs are well-suited to capture such complex relationships as they can incorporate the structural information of the graph along with node and edge features to make predictions. Nodes represent diseases and genes, whereas edges represent known associations between them. Node features can be defined based on various characteristics of diseases and genes, such as Gene Ontology (GO) terms, pathways, and gene expression profiles. Edge features can be defined based on the strength of the association be-tween diseases and genes. The presented research shows a novel pipeline which integrating attention mechanisms into GNNs enhancing model's ability to focus on relevant gene-disease associations, and improving prediction accuracy. Experimental results exhibit a peak accuracy of 80%, highlighting the effectiveness and precision of the approach. This notable accuracy emphasizes the reliability and real-world relevance of the findings for predicting new edges between diseases and genes.}, keywords = {Accuracy, Attention mechanisms, Diseases, Graph neural networks, Ontologies, Pipelines, Predictive models, Proteins, Reliability, Social networking (online)}, pubstate = {published}, tppubtype = {proceedings} } In recent years, Graph Neural Networks (GNNs) have shown great success in predicting missing edges in complex networks such as social networks, protein-protein interaction networks, and knowledge graphs, among others. One of the areas where GNNs may be applied is in predicting new edges between diseases and genes. The relationship between diseases and genes is complex, with a single gene being associated with multiple diseases, and multiple genes being associated with a single disease. GNNs are well-suited to capture such complex relationships as they can incorporate the structural information of the graph along with node and edge features to make predictions. Nodes represent diseases and genes, whereas edges represent known associations between them. Node features can be defined based on various characteristics of diseases and genes, such as Gene Ontology (GO) terms, pathways, and gene expression profiles. Edge features can be defined based on the strength of the association be-tween diseases and genes. The presented research shows a novel pipeline which integrating attention mechanisms into GNNs enhancing model's ability to focus on relevant gene-disease associations, and improving prediction accuracy. Experimental results exhibit a peak accuracy of 80%, highlighting the effectiveness and precision of the approach. This notable accuracy emphasizes the reliability and real-world relevance of the findings for predicting new edges between diseases and genes. |