nempi

nempi infers unobserved perturbations and augments observed ones from incomplete perturbation profiles and differential gene expression in log odds.

Key Features:

• Iterative network-perturbation inference: Alternates between inferring a network from perturbations and inferring perturbations from the network.
• Nested Effects Models: Uses Nested Effects Models for network inference.
• Perturbation profile completion: Handles incomplete perturbation profiles as input.
• Log odds expression input: Processes differential gene expression data in log odds format.

Scientific Applications:

• Cancer driver gene analysis: Analyzes cancer driver genes and their patient-specific perturbation profiles by combining genetic aberrations with gene expression data.
• Perturbation prediction in cancer samples: Predicts perturbations in breast cancer samples from The Cancer Genome Atlas.
• CRISPR perturbation screen validation: Validates perturbation predictions against CRISPR perturbation screens.
• Simulation-based validation: Predicts perturbations in simulation studies.

Methodology:

The method combines genetic aberrations with gene expression data in a causal network derived across patients. Inference proceeds iteratively by inferring a network from perturbations and inferring perturbations from the network, with network inference performed by Nested Effects Models.