ppiPre

ppiPre predicts protein–protein interactions by integrating Gene Ontology (GO) semantic similarities, Kyoto Encyclopedia of Genes and Genomes (KEGG) co-pathway similarity, and topology-based similarities to improve PPI inference across multiple species.

Key Features:

• Heterogeneous Feature Integration: Utilizes three GO-based semantic similarities, one KEGG-based co-pathway similarity, and three topology-based similarities to capture diverse biological signals for PPI prediction.
• Species Support: Supports up to twenty different species for cross-species PPI prediction studies.
• Performance Analysis: Evaluation on binary and co-complex gold-standard yeast PPI datasets reveals significant differences in predictive abilities across features and dataset types, indicating sensitivity to PPI kind and size.
• GO Aspect Utilization: Analysis indicates different GO aspects should be applied to distinct PPI data types, with combining all three GO aspects often improving results and topology-based features alone performing well for co-complex PPIs.

Scientific Applications:

• Systems Biology: Facilitates hypothesis generation of interaction networks and analysis of protein functional relationships.
• Drug Discovery: Aids identification of potential therapeutic targets by predicting novel PPIs.
• Functional Genomics: Assists exploration of protein functional implications within cellular contexts through predicted interactions.

Methodology:

Integration of three GO-based semantic similarities, one KEGG-based co-pathway similarity, and three topology-based similarities for PPI prediction.