BioPathNet: Enhancing Link Prediction in Biomedical Knowledge Graphs through Path Representation Learning DOI
Annalisa Marsico, Yue Hu,

Svitlana Oleshko

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 18, 2024

Abstract Understanding complex interactions in biomedical networks is crucial for advancements biomedicine, but traditional link prediction (LP) methods are limited capturing this complexity. Representation-based learning techniques improve accuracy by mapping nodes to low-dimensional embeddings, yet they often struggle with interpretability and scalability. We present BioPathNet, a novel graph neural network framework based on the Neural Bellman-Ford Network (NBFNet), addressing these limitations through path-based reasoning LP knowledge graphs. Unlike node-embedding frameworks, BioPathNet learns representations between node pairs considering all relations along paths, enhancing interpretability. This allows visualization of influential paths facilitates biological validation. leverages background regulatory (BRG) enhanced message passing uses stringent negative sampling precision. In evaluations across various tasks, such as gene function annotation, drug-disease indication, synthetic lethality, lncRNA-mRNA interaction prediction, consistently outperformed shallow embedding methods, relational task-specific state-of-the-art demonstrating robust performance versatility. Our study predicts drug indications diseases like acute lymphoblastic leukemia (ALL) Alzheimer’s, validated medical experts clinical trials. also identified new lethality involving lncRNAs target genes, confirmed literature reviews. BioPathNet's will enable researchers trace gain molecular insights, making it valuable tool discovery, personalized medicine biology general.

Language: Английский

Path-based reasoning in biomedical knowledge graphs DOI Creative Commons
Yue Hu,

Svitlana Oleshko,

Samuele Firmani

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: June 18, 2024

Abstract Understanding complex interactions in biomedical networks is crucial for advancements biomedicine, but traditional link prediction (LP) methods are limited capturing this complexity. Representation-based learning techniques improve accuracy by mapping nodes to low-dimensional embeddings, yet they often struggle with interpretability and scalability. We present BioPathNet, a novel graph neural network framework based on the Neural Bellman-Ford Network (NBFNet), addressing these limitations through path-based reasoning LP knowledge graphs. Unlike node-embedding frameworks, BioPathNet learns representations between node pairs considering all relations along paths, enhancing interpretability. This allows visualization of influential paths facilitates biological validation. leverages background regulatory (BRG) enhanced message passing uses stringent negative sampling precision. In evaluations across various tasks, such as gene function annotation, drug-disease indication, synthetic lethality, lncRNA-mRNA interaction prediction, consistently outperformed shallow embedding methods, relational task-specific state-of-the-art demonstrating robust performance versatility. Our study predicts drug indications diseases like acute lymphoblastic leukemia (ALL) Alzheimer’s, validated medical experts clinical trials. also identified new lethality involving lncRNAs target genes, confirmed literature reviews. BioPathNet’s will enable researchers trace gain molecular insights, making it valuable tool discovery, personalized medicine biology general.

Language: Английский

Citations

0
BioPathNet: Enhancing Link Prediction in Biomedical Knowledge Graphs through Path Representation Learning DOI
Annalisa Marsico, Yue Hu,

Svitlana Oleshko

et al.

Research Square (Research Square), Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 18, 2024

Abstract Understanding complex interactions in biomedical networks is crucial for advancements biomedicine, but traditional link prediction (LP) methods are limited capturing this complexity. Representation-based learning techniques improve accuracy by mapping nodes to low-dimensional embeddings, yet they often struggle with interpretability and scalability. We present BioPathNet, a novel graph neural network framework based on the Neural Bellman-Ford Network (NBFNet), addressing these limitations through path-based reasoning LP knowledge graphs. Unlike node-embedding frameworks, BioPathNet learns representations between node pairs considering all relations along paths, enhancing interpretability. This allows visualization of influential paths facilitates biological validation. leverages background regulatory (BRG) enhanced message passing uses stringent negative sampling precision. In evaluations across various tasks, such as gene function annotation, drug-disease indication, synthetic lethality, lncRNA-mRNA interaction prediction, consistently outperformed shallow embedding methods, relational task-specific state-of-the-art demonstrating robust performance versatility. Our study predicts drug indications diseases like acute lymphoblastic leukemia (ALL) Alzheimer’s, validated medical experts clinical trials. also identified new lethality involving lncRNAs target genes, confirmed literature reviews. BioPathNet's will enable researchers trace gain molecular insights, making it valuable tool discovery, personalized medicine biology general.

Language: Английский

Citations

0