Predicting Microbe-Disease Association Based on Multiple Similarities and LINE Algorithm.

Area of Science:

• Microbiology
• Computational Biology
• Bioinformatics

Background:

• Microbes significantly impact human health and disease development.
• Understanding microbe-disease associations is crucial for disease diagnosis and treatment.
• Existing methods for inferring these associations have limitations.

Purpose of the Study:

• To develop a novel computational model, MSLINE, for inferring potential microbe-disease associations.
• To integrate multiple similarity measures and network embedding techniques for improved prediction accuracy.
• To enhance the understanding of the human microbiome's role in disease.

Main Methods:

• Constructed a microbe-disease heterogeneous network (MDHN) using known associations and various similarity metrics.
• Applied random walk and Large-scale Information Network Embedding (LINE) to learn network structure.
• Scored potential microbe-disease associations based on learned network information.

Main Results:

• MSLINE demonstrated superior performance compared to existing methods in Leave-one-out and 5-fold cross-validation.
• Case studies confirmed MSLINE's efficiency in predicting novel microbe-disease associations.
• The model successfully integrated diverse data sources for robust association inference.

Conclusions:

• MSLINE is an effective computational tool for predicting microbe-disease associations.
• The integration of multiple similarities and network embedding enhances prediction accuracy.
• This approach offers valuable insights for microbiome-based disease research and diagnostics.