Reaction Networks, Autocatalytic Sets, and Regulatory Architectures: How Stoichiometric Unification, Sampling Bias, Spatial Context, and Controllability Constraints Jointly Shape a Candidate Framework for Biological Network Design Principles

Abstract

This version corrects one citation error found by an automated check and confirmed by hand: a paper on self-regulatory communication in evolved neural agents, listed among those considered and not included, was cited under the identifier 2602.02840, which belongs to an unrelated paper; the correct identifier is arXiv:2606.02840. No claim of the synthesis changes. Dashes were also normalised. This version has not had a full claim-by-claim audit. Biological networks, from the prebiotic chemistry that may have seeded life, to the gene regulatory circuits controlling cell identity, to the ecological webs structuring communities, are often treated as separate objects of study with distinct formalisms. This synthesis proposes a candidate reading, explicitly heuristic rather than derivational, that a set of shared design constraints recurs across these scales: (i) the structural conditions under which a network becomes self-amplifying, (ii) the sampling and representational biases that distort our inference of network properties, (iii) the role of spatial context in determining whether network-level cooperation or regulation is stable, and (iv) the controllability cost of returning a network to a functional state. Drawing on six papers from q-bio.MN, q-bio.GN, q-bio.PE, and q-bio.BM, we synthesize findings spanning autocatalytic set theory arXiv:2605.25523, Boolean network sampling bias arXiv:2606.05196, tissue graph counterfactuals arXiv:2606.08493, eco-evolutionary cooperation dynamics arXiv:2606.03071, and control-theoretic aging frameworks arXiv:2605.16781. An information-theoretic regulatory blueprint paper arXiv:2605.19071 is retained as a suggestive parallel in the weakly-connected addendum; its bridge to the main synthesis is vocabulary-level rather than mechanistic, and we do not treat it as a load-bearing claim. We argue that each primary domain independently surfaces the same structural tension: network topology and the functions we assign to it are jointly underdetermined without explicit accounting for spatial context, representational completeness, and dynamical reachability. The central falsification path is specific: if function-uniform sampling of Boolean networks arXiv:2606.05196 and spatial disentanglement in tissue graphs arXiv:2606.08493 are each correcting for the same class of context-collapse bias, then applying function-uniform priors to spatially-embedded regulatory network ensembles should shift attractor-structure predictions in a direction consistent with the tissue-level counterfactuals. Failure of that prediction would falsify the cross-domain bridge proposed here. Authorship: Saluca Agentic AI Research Team (Saluca LLC). AI-drafted synthesis from an arXiv preprint corpus, originally drafted 2026-06-14, produced under the direction of Cristian Ruvalcaba, the accountable human author. Not peer-reviewed. AI disclosure. This work was produced with an agentic AI research apparatus operated by Saluca Labs. The apparatus drafted, searched and analysed under direction. Cristian Ruvalcaba is the human author and is accountable for the content. No AI system is listed as an author or contributor, because authorship entails accountability that a model cannot hold; this disclosure is the credit, and it is deliberately the whole of it.