aWCluster

aWCluster integrates network-based multiomics profiles using the Wasserstein distance (optimal mass transport) to cluster samples and identify molecular subtypes.

Key Features:

• Network-Based Integration: Integrates mRNA expression, DNA copy number alterations, and DNA methylation data through interaction networks that consider relationships between genes and their neighbors.
• Wasserstein Distance Clustering: Uses the Wasserstein distance from optimal mass transport theory to measure dissimilarity between multiomics profiles for clustering.
• Subtype Characterization: Identifies cancer subtypes by analyzing concordant effects across different omics layers and delineates classes with distinct survival outcomes in TCGA datasets.
• Biomarker Discovery: Focuses on genes with concordant multiomics measurements within their interaction networks to enhance identification of potential biomarkers and novel subtypes.
• Gene Ontology Enrichment Analysis: Performs gene ontology enrichment to highlight significant biological processes and factors, including tumor hypoxia and the transcription factor ETS1 in breast cancer studies.

Scientific Applications:

• Cancer subtype discovery and survival stratification: Applied to TCGA cohorts including breast carcinoma, bladder carcinoma, colorectal adenocarcinoma, renal carcinoma, lung non-small cell adenocarcinoma, and endometrial carcinoma to delineate classes with distinct survival outcomes.
• Biomarker and pathway analysis: Enables identification of candidate biomarkers and enriched biological processes (for example, tumor hypoxia and ETS1-associated pathways) by integrating multi-layer omics signals within interaction networks.

Methodology:

Integrates mRNA expression, DNA copy number alterations, and DNA methylation via gene interaction networks considering gene neighbors; computes Wasserstein distance (optimal mass transport) between network-based multiomics profiles for clustering; identifies genes with concordant multiomics measurements within networks and applies gene ontology enrichment analysis.