NEArender

NEArender transforms raw 'omics' feature matrices into network enrichment analysis (NEA)-based pathway score matrices to improve statistical power and biological relevance in pathway enrichment and phenotype modeling, and is implemented as an R package.

Key Features:

• Parametric null estimation: Employs a parametric estimation of the null binomial distribution to evaluate network enrichment scores.
• Matrix rendering: Converts raw gene profile matrices (N genes x N samples) into pathway enrichment score matrices (N pathways x N samples) using NEA.
• Increased statistical power: Summarizes individual genes into functionally annotated gene sets to enhance statistical power relative to per-gene differential expression and gene set enrichment analyses.
• Comprehensive functionality: Provides functions for preparing input data, modeling null distributions, and evaluating alternative versions of the global network.
• Integration for modeling: Produces pathway score matrices intended for use as features in phenotype modeling and disease outcome prediction pipelines.

Scientific Applications:

• Molecular landscape exploration: Enables characterization of sample-level pathway activity across functionally annotated gene sets.
• Predictive modeling: Supplies NEA-based pathway scores as input features for phenotype and disease outcome prediction.
• Mechanism and target identification: Facilitates interpretation of complex biological processes, disease mechanisms, and potential therapeutic targets via pathway-centric analysis.

Methodology:

Applies network enrichment analysis (NEA) with a parametric estimation of the null binomial distribution to render N genes x N samples matrices into N pathways x N samples pathway score matrices; includes routines for input preparation, null-distribution modeling, and evaluation of alternative global-network versions.