m6Adecom

m6Adecom applies graph-regularized non-negative matrix factorization to analyze N6-methyladenosine (m6A) methylation profiles and integrate RNA binding protein (RBP) preferences for functional interpretation.

Key Features:

• Graph-regularized NMF (GNMF): Employs graph-regularized non-negative matrix factorization (GNMF) that incorporates RNA binding protein (RBP) binding preferences as graph constraints to decompose m6A profile matrices.
• Functional context-aware analysis: Integrates RBP binding data to capture functional distinctions among groups of m6A sites linked to biological pathways and disease genes.
• Correlation and enrichment analyses: Supports correlation analyses and gene set enrichment analysis (GSEA) using high-scoring m6A sites obtained from GNMF decomposition.

Scientific Applications:

• Correlation and enrichment analysis: Identifies associations between m6A methylation patterns and biological functions or diseases through correlation analyses.
• Gene Set Enrichment Analysis (GSEA): Performs GSEA on high-scoring m6A sites derived from GNMF to reveal enriched gene sets associated with specific m6A modifications.
• Pathway and disease gene analysis: Detects pathways and disease-associated genes by distinguishing functional characteristics of different m6A site groups.

Methodology:

Applies graph-regularized non-negative matrix factorization (GNMF) to decompose m6A profile matrices with a graph constraint term derived from RBP binding preferences, and uses the resulting factorization for downstream correlation analyses and gene set enrichment.