HumanNet

HumanNet v3 constructs an integrated human gene network covering 99.8% of human protein-coding genes to support network-medicine analyses and identification of disease-associated genes.

Key Features:

• Integrated Network Model: A three-tier model comprising HumanNet-PI (Protein-Protein Physical Interaction Network) for protein physical interactions, HumanNet-FN (Functional Gene Network) for gene function and phenotype associations, and HumanNet-XC (Extended Functional Network by Co-citation) incorporating co-citation data to extend functional associations.
• Data Integration and Expansion: Integration of numerous large-scale protein-protein interaction datasets from public repositories and expanded functional annotations and research-paper-derived associations for gene-phenotype links.
• Improved Network Inference Algorithms: Refined network inference algorithms that enhance the precision and reliability of disease gene predictions compared to previous versions.

Scientific Applications:

• Disease Gene Prediction: Prioritization and prediction of genes associated with complex human diseases using the integrated network model.
• COVID-19 Research: Identification of host genes potentially linked to COVID-19 through network-based analyses incorporating host interaction and functional associations.

Methodology:

Integration of protein-protein interaction datasets, functional annotations including gene-phenotype associations, and co-citation data, followed by application of refined network inference algorithms to construct the three-tier networks.