Search bioRxiv⌕ Search

Biology subjects

Janardhanan, S.

Publications and source records attributed to Janardhanan, S..

2 recordsLinked to original sources

Certified Steady-State Parameter-Interval Design for Uncertain Biomolecular Models using a Global Shaving Contractor

The design of parameter intervals that provably enforce steady-state specifications in biomolecular circuits is challenging due to nonlinear reaction kinetics, parametric uncertainty, and the under-determined nature of steady-state constraints. Most validated approaches either rely on recursive subdivision (set inversion) or may stall due to dependency effects when applied directly in parameter space, limiting scalability in moderate to high dimensions. This paper introduces a global shaving contractor that contracts an initial parameter box by repeatedly applying certified interval-exclusion tests against a prescribed steady-state set. The proposed procedure returns a guaranteed outer enclosure of the feasible parameter set and provides finite-termination guarantees, along with worst-case bounds on the number of inclusion-function evaluations. Case studies spanning low-dimensional motifs and a sixteen-parameter integral-feedback model, including bistability specifications for a CRISPRi toggle switch, demonstrate substantial contraction of design domains without subdivision. The resulting certificates support uncertainty-aware circuit tuning, rigorous parameter screening, and robust design workflows in systems & synthetic biology and related nonlinear dynamical-system design problems.

synthetic biology↗

Design of Parameter Intervals to Meet Steady-State Specifications in Biomolecular Circuits using Interval Analysis

It is often desired, in the analysis or the design of biomolecular circuits, to search for parameters that satisfy certain specifications, such as on the number of the steady states or on their magnitudes. These problems are challenging because of the presence of multiple parameters, the nonlinear mapping between the parameters and the steady states, as well as the specifications themselves, such as when multiple steady states are desired. Typically, exact analytical solutions are limited and numerical approaches, if they converge, may not capture all solutions. We used Interval Analysis versions of Bisection and Constraint Propagation to obtain rigorous and guaranteed estimates of circuit parameters for the design problem. We established criteria that rule out the existence of design solutions in a given parameter space. We presented algorithms to rigorously bound all solutions and developed variants that enclosed the solutions as accurately as required. These theoretical results were illustrated on benchmark feedback and feedforward circuits. These results should aid in the analysis and design of biomolecular circuits as well as in other contexts with similar modelling frameworks. The rigorous nature of the results may be particularly useful for resource optimization and safety criticality in synthetic biology applications, such as in therapeutic applications and drug delivery.

synthetic biology↗