Optimization-based synthesis of stochastic biocircuits with statistical specifications.

Area of Science:

• Synthetic biology
• Computational biology
• Biophysics

Background:

• Model-guided design is standard in synthetic biology.
• Stochasticity in biomolecular reactions is often ignored, leading to deviations from predictions.
• Cell-cell heterogeneity complicates biocircuit behavior and necessitates design-build-test cycles.

Purpose of the Study:

• To present a computational framework for designing stochastic biocircuits.
• To systematically specify and control intrinsic noise levels using statistical metrics.
• To design highly heterogeneous biocircuits based on statistical specifications.

Main Methods:

• Utilizing descriptive statistics of population distributions as a specification language.
• Developing an optimization-based computational tool to explore parameter configurations.
• Employing sensitivity analysis for robustness against extrinsic perturbations.
• Formulating design tools with convex optimization for efficient quantification.

Main Results:

• Demonstrated the design of a stochastic negative feedback biocircuit meeting multiple statistical constraints.
• Provided an in-depth analysis of noise propagation and regulation in negative feedback pathways.
• Successfully integrated statistical specifications into biocircuit design.

Conclusions:

• The developed framework enhances the design of stochastic biocircuits by controlling intrinsic noise.
• This approach improves predictability and reduces the need for extensive design iterations.
• Offers a rigorous and efficient method for quantifying statistics in biocircuit design.