Using random perturbations to infer the structure of feedback control in gene expression.

Area of Science:

• Systems biology
• Molecular systems biology
• Computational biology

Background:

• Cellular feedback mechanisms are crucial for biological regulation.
• Inferring feedback typically relies on observing cellular responses to specific experimental perturbations.
• Existing frameworks like modular response analysis link perturbations to feedback sensitivities but may not cover complex, multi-module drug effects.

Purpose of the Study:

• To analyze gene expression model responses to simultaneous, random perturbations across multiple modules.
• To investigate the applicability of deterministic analysis in inferring feedback regulation under complex perturbation scenarios.
• To identify and explain discrepancies between deterministic predictions and stochastic gene expression dynamics.

Main Methods:

• Analytical derivation of feedback sensitivities from infinitesimal random perturbations in the deterministic regime.
• Numerical extension of deterministic analysis to finite perturbations in stochastic gene expression models.
• Sampling of simple stochastic gene expression models to identify deviations from deterministic predictions.

Main Results:

• Deterministic analysis can infer gene expression feedback if perturbation effects are statistically independent between modules.
• Stochastic gene expression models exhibit violations of deterministic bounds, even with infinitesimal perturbations.
• These violations are attributed to the inherent stochastic dynamics of biochemical feedback circuits.

Conclusions:

• Deterministic analyses may inaccurately predict the response of cellular averages to perturbations, even within the linear response regime.
• Stochastic effects are critical for accurately modeling biochemical feedback circuits and their responses.
• Understanding these limitations is essential for interpreting large-scale drug perturbation studies.