Spatial Compilation of Stochastic Chemical Trajectories
Year: 2026
Type: preprint
Abstract: Can a well-mixed reaction specification be realized by particles that move, collide and sometimes reject a proposed product? This paper constructs such a spatial realization for finite conservative, reversible stochastic reaction networks, with variable reaction order and distinct null and parallel labels. The compiler emits elementary local reaction channels, an integer payload decoder, strictly positive task-dependent particle radii, constant rational coefficients and one known slowdown. Original particles may start at any feasible deterministic placement, including simultaneous wall and pair contacts; all auxiliary populations start at zero. The decoded record approximates the law of the entire timed and labelled target trajectory on a prescribed finite horizon, retaining exact rescaled event-time coordinates, original labels and post-event counts. No initial waiting interval is discarded. Transport may be unknown, species dependent, spatially inhomogeneous and time dependent, within the stated ellipticity and regularity bounds. The same output tolerates a computed box of predictable chemical, drift and correlated covariance faults whose policies may read the full continuous and hidden event history. The reflection law remains the specified nominal block-conormal law. A dilute hard-ball Poincaré inequality supplies the missing many-particle mixing estimate. Its proof uses a small inward buffer, rank separation with disjoint chamber images, and common intermediate paths that move one three-dimensional particle coordinate at a time. Exact forward/reverse placement fluxes preserve nominal spatial detailed balance. Conditional comparison and proposal coupling then retain the actual hidden history while controlling the complete observed path. The resource theorem pays for physical duration, inverse radii, rate and activity reciprocals, precision, every successful or rejected attempt, particle motion and every actual wall and pair push. Bounds are polynomial in a stated numerical charge together with the complete binary input length; dilution, time and precision costs may be severe. Push lengths are geometric quantities, not thermodynamic energy. The input requires a positive conserved integer weight, a reverse-label involution and prescribed rational detailed-balanced activities. The theorem concerns an ideal finite-radius model in a three-dimensional convex polytope. It does not establish a general open-network spatial compiler, a fixed-density theorem, exact equality of path laws, real molecular synthesis, leakage control or practical affordability. The kinetic dependency is the separately proved balanced-conservative Theorem 9.1 of Compiling Stochastic Chemical Trajectories with Elementary Reactions; the general open kinetic theorem has a different scope.
Author: Sandhu, Harjas
Institution: The University of Queensland
Language: English
Topic: DNA and Biological Computing
Subfield: Theoretical Computer Science
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225996978
