APPAGATO

APPAGATO identifies approximate occurrences of query biological networks within larger target networks, accommodating topological differences and mismatches in nodes, edges, and node labels.

Key Features:

• Approximate Querying: Finds approximate matches between a query network and a target network to tolerate noise and inconsistencies in biological data.
• Stochastic and Parallel Processing: Uses a stochastic sampling algorithm combined with parallel processing to improve scalability for large-scale networks.
• Handling of Mismatches: Manages mismatches in nodes, edges, and node labels to produce meaningful partial or inexact matches.
• Performance and Accuracy: Demonstrates improved computational performance and statistically significant accuracy gains relative to existing tools.
• CUDA-C++ Toolkit Integration: Implements GPU-accelerated computation using the CUDA-C++ Toolkit 7.0 framework.

Scientific Applications:

• Protein–Protein Interaction Network Analysis: Identifies and compares protein complexes across different species and tissues within protein–protein interaction networks.
• Gene Ontology–Based Annotation: Utilizes synthetic and real Gene Ontology terms for annotation to infer functional similarities and evolutionary relationships.

Methodology:

Employs a stochastic algorithm that randomly samples potential matches in the target network and uses parallel processing (implemented with CUDA-C++ Toolkit 7.0) to accelerate computations and statistically validate results across multiple iterations.