sub-SPIA

sub-SPIA identifies signaling subpathways by integrating subpathway analysis into signaling-pathway impact analysis (SPIA) to improve detection of pathways perturbed by differentially expressed genes, particularly in cancer.

Key Features:

• Combination of analyses: Combines classical enrichment analysis with perturbation-based signaling-pathway impact analysis (SPIA) to assess pathway impact and perturbation.
• Subpathway integration: Integrates subpathway analysis into SPIA to capture localized regions of differential expression within pathways.
• MST-based subpathway definition: Defines subpathways using minimal-spanning-tree (MST) structures to flexibly capture complex subpathway architectures.
• Improved resolution and reduced overlap: Reduces overlapping subpathways and improves resolution compared with prior k-clique-based subpathway definitions.
• Cancer dataset application: Applied to colorectal cancer and lung cancer gene expression datasets to identify significant pathways missed by other methods.
• KEGG network analysis: Analyzes the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway network to evaluate average degree and connectivity of identified pathways.
• Biological insight: Highlights pathways with larger average degree and closer connectivity as potential conduits for abnormal signal propagation in disease processes.

Scientific Applications:

• Cancer pathway identification: Identification of cancer-related signaling subpathways in colorectal cancer and lung cancer datasets.
• Mechanistic prioritization: Prioritization of pathways that may mediate abnormal signal propagation driving specific disease processes.
• Comparative network analysis: Comparative evaluation of pathway connectivity and degree within KEGG to distinguish pathways implicated by sub-SPIA from those found by other methods.

Methodology:

Combines classical enrichment analysis with perturbation calculations from SPIA; integrates subpathway analysis defined by minimal-spanning-tree (MST) structures (contrasted with prior k-clique definitions); applied to colorectal cancer and lung cancer gene expression datasets; performs network analysis on the KEGG pathway network to compute average degree and connectivity.