A Constructibility Theorem for Protein-Only Turing-Complete Computation: Sequential-Head Reduction via NAND-Cooperative Networks and Protein-Templated Memory Primitives

We construct a formal computational model — the Protein Computational Substrate (PCS) — whose primitives are restricted to protein-mediated operations on a passive DNA medium. We prove two principal results: (i) any PCS satisfying five explicit axioms (NAND universality, head-local write, head-local read, unbounded tape extension, and head coherence) simulates an arbitrary Turing machine via the standard sequential-head reduction, and (ii) head coherence (A5), under a precise "independent-position" formalization of its negation, acts as a structural separator between the FSA and TM computational classes within the framework — relaxing A5 in this sense collapses the PCS to a finite-state automaton even when the other primitives are individually preserved. The first result is a specialization of the chemical-Turing universality lineage (Bennett 1982; Hjelmfelt–Weinberger–Ross 1991, 1992; Rothemund 1995; Magnasco 1997; Shapiro 2012; Soloveichik 2008, 2010; among others) to a strictly protein-only substrate. The second result provides, to our knowledge, one of the first explicit axiomatic separations of this kind within a biomolecular computation framework, and is consistent with the observation that prior in-cell recombinase state machines (Roquet et al. 2016; Benenson–Shapiro 2001) realize FSAs. The overall theorem is a constructibility result in the spirit of Turing's 1936 construction: it identifies a minimal set of protein-only primitives sufficient for universal computation without claiming that any extant organism realizes such a configuration. Each axiom is shown to be ingredient-consistent with known classes of proteins — the necessary component primitives are independently established (cooperative allosteric enzymes for NAND [1, 2]; the Drt3b component of the DRT3 antiphage defense system, in which a protein-templated poly(AC) DNA strand is synthesized from active-site residues of the protein itself [3], as the existence proof for protein-templated DNA write; sequence-specific DNA-binding proteins for read; DNA polymerase for tape extension) — but no claim is made that any single integrated system satisfying all five axioms simultaneously has been constructed or observed. The integration of these primitives into a single operating substrate is open and forms the central question for synthetic biology that the construction makes explicit. We separate the mathematical content of the theorem (rigorous) from the biological-instantiation question (open). The memory-primitive lemma is presented as an abstract specification, with concrete realizations deferred to companion work.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-05-31
DOI
https://doi.org/10.5281/zenodo.20375110
Primary Topic
DNA and Biological Computing
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

A Constructibility Theorem for Protein-Only Turing-Complete Computation: Sequential-Head Reduction via NAND-Cooperative Networks and Protein-Templated Memory Primitives

Eunjoon Jang
Zenodo (CERN European Organization for Nuclear Research)
DNA and Biological Computing
preprint

A Constructibility Theorem for Protein-Only Turing-Complete Computation: Sequential-Head Reduction via NAND-Cooperative Networks and Protein-Templated Memory Primitives

Eunjoon Jang
preprint en

Abstract

We construct a formal computational model — the Protein Computational Substrate (PCS) — whose primitives are restricted to protein-mediated operations on a passive DNA medium. We prove two principal results: (i) any PCS satisfying five explicit axioms (NAND universality, head-local write, head-local read, unbounded tape extension, and head coherence) simulates an arbitrary Turing machine via the standard sequential-head reduction, and (ii) head coherence (A5), under a precise "independent-position" formalization of its negation, acts as a structural separator between the FSA and TM computational classes within the framework — relaxing A5 in this sense collapses the PCS to a finite-state automaton even when the other primitives are individually preserved. The first result is a specialization of the chemical-Turing universality lineage (Bennett 1982; Hjelmfelt–Weinberger–Ross 1991, 1992; Rothemund 1995; Magnasco 1997; Shapiro 2012; Soloveichik 2008, 2010; among others) to a strictly protein-only substrate. The second result provides, to our knowledge, one of the first explicit axiomatic separations of this kind within a biomolecular computation framework, and is consistent with the observation that prior in-cell recombinase state machines (Roquet et al. 2016; Benenson–Shapiro 2001) realize FSAs. The overall theorem is a constructibility result in the spirit of Turing's 1936 construction: it identifies a minimal set of protein-only primitives sufficient for universal computation without claiming that any extant organism realizes such a configuration. Each axiom is shown to be ingredient-consistent with known classes of proteins — the necessary component primitives are independently established (cooperative allosteric enzymes for NAND [1, 2]; the Drt3b component of the DRT3 antiphage defense system, in which a protein-templated poly(AC) DNA strand is synthesized from active-site residues of the protein itself [3], as the existence proof for protein-templated DNA write; sequence-specific DNA-binding proteins for read; DNA polymerase for tape extension) — but no claim is made that any single integrated system satisfying all five axioms simultaneously has been constructed or observed. The integration of these primitives into a single operating substrate is open and forms the central question for synthetic biology that the construction makes explicit. We separate the mathematical content of the theorem (rigorous) from the biological-instantiation question (open). The memory-primitive lemma is presented as an abstract specification, with concrete realizations deferred to companion work.

Zenodo (CERN European Organization for Nuclear Research)
Inha University (KR)
DNA and Biological Computing
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.