response logic

response logic infers directed regulatory networks from perturbation-response data to identify signal propagation and regulatory interactions in molecular and cellular systems.

Key Features:

• Minimal Presumptions: Requires only the experimental observability of whether system components respond to perturbations, enabling application across diverse biological datasets.
• Logic Programming Approach: Utilizes a logic programming framework, including Answer Set Programming (ASP), to efficiently infer directed networks and manage combinatorial complexity for networks of hundreds of nodes.
• Robustness to Data Imperfections: Handles noisy, heterogeneous, and missing data to maintain reliable network inference under imperfect experimental conditions.
• Integration of Prior Knowledge: Incorporates prior network knowledge and explicit constraints such as sparsity directly into the inference process.
• Benchmarking Success: Demonstrated improved performance in systematic benchmarking on KEGG pathways and in challenges such as DREAM3 and DREAM4.

Scientific Applications:

• Pathway-level Network Inference: Infers regulatory interactions in signalling pathways such as PI3K and MAPK from perturbation datasets.
• Cancer Model Analysis: Generates network hypotheses in isogenic models of a colon cancer cell line to explain differential sensitivities to targeted inhibitors associated with different PI3K mutants.

Methodology:

Observes component responses to perturbations and identifies directed networks that best explain observed signal propagation using a logic programming approach implemented with Answer Set Programming (ASP) and Python.