Are Quasi-Steady-State Approximated Models Suitable for Quantifying Intrinsic Noise Accurately?

Area of Science:

• Systems Biology
• Computational Biology
• Molecular Systems Engineering

Background:

• Gene regulatory networks (GRN) are often simplified using quasi-steady-state approximation (QSSA) to reduce computational time for intrinsic noise quantification via Gillespie stochastic simulation algorithm (SSA).
• The accuracy of QSSA in measuring intrinsic noise compared to detailed mechanistic models remains a critical question in systems biology.

Purpose of the Study:

• To evaluate the accuracy of stochastic QSSA models in quantifying intrinsic noise for gene regulatory networks (GRN) compared to detailed mechanistic models.
• To determine the factors influencing the accuracy of QSSA in noise quantification for GRNs.

Main Methods:

• Construction of mechanistic and QSSA models for frequently observed GRNs exhibiting switching behavior.
• Performance of stochastic simulations using both model types.
• Analysis of intrinsic noise quantification accuracy at both mRNA and protein levels.

Main Results:

• Stochastic QSSA model accuracy critically depends on the absolute values of mRNA and protein half-lives, not their abundance levels.
• QSSA provides greater accuracy for protein noise quantification across a wider range of half-life values compared to mRNA.
• Satisfactory accuracy for mRNA noise quantification is achieved only for limited combinations of absolute half-life values.

Conclusions:

• Stochastic QSSA models can be a reliable choice for evaluating intrinsic noise in other GRNs.
• Model selection should be guided by experimentally determined mRNA and protein half-life values.
• QSSA offers a computationally efficient alternative for GRN noise analysis when half-lives are considered.