A Pathway Analysis Approach Using Petri Net.

Area of Science:

• Computational Biology
• Systems Biology
• Bioinformatics

Background:

• Cell signaling pathway perturbations are linked to various diseases.
• Accurate identification of these perturbed pathways is crucial for understanding disease mechanisms.
• Existing pathway analysis methods, particularly topology-based ones, have limitations due to simplified models.

Purpose of the Study:

• To introduce a novel Pathway Analysis method using Petri nets (PAPet) for modeling and analyzing cell signaling pathways.
• To develop an algorithm for measuring pathway perturbation under specific conditions.
• To overcome the limitations of traditional graph-based models in pathway analysis.

Main Methods:

• Modeling cell signaling pathways using Petri nets.
• Developing a perturbation measurement algorithm for Petri net models.
• Evaluating PAPet using sensitivity, prioritization, mean reciprocal rank, and false-positive rate metrics on 36 real disease datasets.
• Comparing PAPet against five established methods: FoPA, PADOG, GSEA, CePa, and SPIA.
• Applying PAPet to gene expression profiles from normal and Pancreatic Ductal Adenocarcinoma (PDAC) samples.

Main Results:

• PAPet demonstrated a superior balance across sensitivity, prioritization, mean reciprocal rank, and false-positive rate metrics compared to FoPA, PADOG, GSEA, CePa, and SPIA.
• PAPet achieved the highest rank in identifying known PDAC-associated pathways when applied to gene expression data.
• The Petri net modeling approach in PAPet effectively addresses shortcomings of simpler graph-based models.

Conclusions:

• PAPet represents a significant advancement in pathway analysis, offering improved accuracy and robustness.
• The method provides a valuable tool for investigating disease mechanisms through the analysis of perturbed cell signaling pathways.
• PAPet is recommended for its comprehensive performance and applicability in disease research, including cancer studies.