PPIntegrator

PPIntegrator integrates heterogeneous protein-protein interaction (PPI) datasets using semantic web standards to formalize and interlink data into biological knowledge graphs for host–pathogen interaction analysis.

Key Features:

• Semantic Description and Data Integration: Uses ontologies to semantically describe PPI-related data and provide consistent metadata annotation for integration.
• Data Preparation and Triplification: Includes a data preparation module that organizes information from three reference databases and converts data into a structured triple format via triplification.
• Provenance Information and Data Fusion: Captures provenance information during data fusion to ensure traceability of data sources and applied transformations.
• Enrichment Pipeline: Generates, predicts, and validates new potential host–pathogen PPI datasets through transitivity analysis.
• Transitivity Analysis for Host–Pathogen Datasets: Applies transitivity analysis to integrate and compare host–pathogen PPI datasets across multiple bacterial species, identifying indirect interactions.
• Critical Query Support: Supports critical queries for analysis of integrated PPI data to extract hypotheses about protein functions and interactions.

Scientific Applications:

• Protein Function Inference: Infers protein functions from integrated and transitivity-inferred PPI networks.
• Host–Pathogen Interaction Analysis: Integrates and compares host–pathogen PPI datasets across bacterial species to study interaction dynamics.
• Hypothesis Generation and Target Identification: Expands PPI datasets to support generation of interaction hypotheses and identification of candidate therapeutic targets in microbial pathogenesis.
• Knowledge Graph Construction and Interoperability: Formalizes and interlinks PPI data into biological knowledge graphs to enable interoperable analyses across databases.

Methodology:

PPIntegrator semantically describes PPI data with ontologies, prepares and organizes information from three reference databases, performs triplification to convert data into triple format, applies data fusion while capturing provenance, runs an enrichment pipeline that generates/predicts/validates new host–pathogen PPIs via transitivity analysis, integrates and compares host–pathogen datasets from multiple bacterial species, and supports critical queries on the integrated graph.