lpNet

lpNet infers signaling and gene regulatory networks from high-throughput perturbation and expression data to reconstruct interactions among coding and non-coding mRNA and proteins.

Key Features:

• Data Input Flexibility: Accepts perturbation and non-perturbation data, including steady-state and time-series datasets, for inference from diverse experimental setups.
• Linear Programming Approach: Uses linear programming (LP) to reconstruct interactions among biological factors.
• Efficiency and Speed: Demonstrated efficient reconstruction of underlying networks on simulated and real datasets.
• Validation Techniques: Enables parameter identification and validation via leave-one-out cross-validation and stratified n-fold cross-validation.

Scientific Applications:

• Signaling network inference: Infers signaling interactions within cellular networks.
• Gene regulatory network inference: Reconstructs gene regulatory interactions that govern mRNA and protein levels.
• Elucidation of regulatory mechanisms and disease insights: Provides insights into cellular regulatory interactions and disease mechanisms from high-throughput data.

Methodology:

Linear programming-based reconstruction from combined perturbation/non-perturbation and steady-state/time-series data; parameter identification using leave-one-out or stratified n-fold cross-validation; validation on simulated and real datasets.