probabilistic graphlets

probabilistic graphlets analyze local topological patterns in probabilistic molecular interaction networks by incorporating edge probabilities to characterize biological functions.

Key Features:

• Probabilistic Analysis: Incorporates edge weights as probabilities to examine the likelihood and strength of molecular interactions.
• Enhanced Sensitivity: Maintains probabilistic data to improve sensitivity in distinguishing different network structures compared to unweighted graphlet methods.
• Robust Biological Information Capture: Preserves interaction probability information in weighted networks to retain biological signal lost by thresholding to binary edges.

Scientific Applications:

• Condition-specific function identification: Identifies condition- or perturbation-specific biological functions within molecular interaction networks by analyzing probabilistic interactions.

Methodology:

Generates synthetic networks from established random network models with various edge probability distributions to benchmark probabilistic graphlets against unweighted counterparts, and applies the method to real-world molecular interaction networks modeled as weighted graphs with probabilistic edges.