GRaNPA

GRaNPA evaluates the performance and biological relevance of enhancer-mediated gene regulatory networks by predicting cell-type-specific differential gene expression from chromatin accessibility and RNA sequencing data.

Key Features:

• Enhancer-Mediated GRN Inference: Builds on the companion tool GRaNIE to construct enhancer-mediated gene regulatory networks by analyzing covariation between chromatin accessibility data and RNA sequencing across samples such as individuals or cell types.
• Cell-Type-Specific Analysis: Focuses on cell-type-specific gene regulation to capture interactions between transcription factors and enhancers that influence gene expression in different cellular contexts.
• Performance Assessment: Provides an unbiased assessment of inferred GRNs by evaluating their ability to predict differential gene expression across cell types in the absence of a complete ground truth.
• Application in Disease and Trait Research: Applies to studies of macrophage responses to infection, cancer progression, and the genetic basis of autoimmune diseases.
• Identification of Key Regulators: Identifies candidate key regulators within GRNs, including implicating the transcription factor PURA in pro-inflammatory macrophage polarization.

Scientific Applications:

• Dissection of Gene Regulatory Mechanisms: Dissects how enhancers and transcription factors contribute to cell-type-specific gene expression and complex trait regulation.
• Immune Response Studies: Investigates regulatory mechanisms underlying macrophage polarization and responses to infection.
• Cancer Research: Explores gene regulatory changes associated with cancer progression.
• Autoimmune Disease Genetics: Evaluates how genetic variants and epigenetic modifications influence gene expression relevant to autoimmune disease etiology.

Methodology:

Operates within a machine learning framework that predicts cell-type-specific differential expression from enhancer-mediated GRNs inferred by GRaNIE using covariation between chromatin accessibility data and RNA sequencing across samples, and assesses GRN performance by testing prediction accuracy of differential gene expression across cell types.