epiNEM

epiNEM extends Nested Effects Models (NEMs) to infer genetic interaction networks, account for double knockouts, and identify modulators of mixed epistasis from high-dimensional gene perturbation readouts.

Key Features:

• Extension of NEMs: Extends the Nested Effects Models framework to model combinatorial genetic perturbations such as double knockouts.
• Handling Mixed Epistasis: Models mixed epistasis by allowing a third gene to modulate interactions between two other genes.
• Logical Functions Integration: Incorporates logical functions to describe interactions between regulators and downstream genes and proteins.
• High-Dimensional Data Analysis: Analyzes high-dimensional gene expression profiles as molecular readouts of combinatorial perturbations.
• Inference of Signaling Pathways and Modulators: Infers network signaling pathways and identifies modulators of genetic interactions, including analyses of deletion mutants of kinases and phosphatases in Saccharomyces cerevisiae.
• Benchmarking and Accuracy: Validated by simulation studies demonstrating high accuracy in recovering correct models of genetic interaction networks.
• Implementation: Implemented as an R package.

Scientific Applications:

• Mapping Functional Redundancies: Dissects functional redundancies within cellular networks by mapping genetic interactions.
• Identifying Modulators: Identifies third-party modulators of gene–gene interactions, demonstrated on deletion mutants of kinases and phosphatases in Saccharomyces cerevisiae.
• Signaling Pathway Reconstruction: Reconstructs signaling pathways from combinatorial perturbation readouts.

Methodology:

Extends the NEM framework to incorporate logical functions describing regulator–target interactions, models combinatorial perturbations including double knockouts from high-dimensional gene expression readouts, and uses simulation studies for benchmarking.