## Efficient Turing-universal computation with DNA polymers (extended abstract)

Citations: | 15 - 2 self |

### BibTeX

@MISC{Qian_efficientturing-universal,

author = {Lulu Qian and David Soloveichik and Erik Winfree},

title = {Efficient Turing-universal computation with DNA polymers (extended abstract)},

year = {}

}

### OpenURL

### Abstract

Abstract. Bennett’s proposed chemical Turing machine is one of the most important thought experiments in the study of the thermodynamics of computation. Yet the sophistication of molecular engineering required to physically construct Bennett’s hypothetical polymer substrate and enzyme has deterred experimental implementations. Here we propose a chemical implementation of stack machines — a Turing-universal model of computation similar to Turing machines — using strand displacement cascades as the underlying chemical primitive. More specifically, the mechanism described herein is the addition and removal of monomers from the end of a polymer, controlled by strand displacement logic. We capture the motivating feature of Bennett’s scheme — that physical reversibility corresponds to logically reversible computation, and arbitrarily little energy per computation step is required. Further, as a method of embedding logic control into chemical and biological systems, polymer-based chemical computation is significantly more efficient than geometry-free chemical reaction networks. 1

### Citations

641 | Molecular computation of solutions to combinatorial problems
- Adleman
- 1994
(Show Context)
Citation Context ...Turing machine has been hampered by our inability, as yet, to engineer molecular machinery to spec. Len Adleman’s laboratory demonstration of a DNA computing paradigm for solving NP-complete problems =-=[1]-=- ignited renewed interest in the molecular implementation of Turing machines. Early theoretical proposals made use of existing enzymes but required a series of laboratory manipulations to step the mol... |

471 |
Logical reversibility of computation
- Bennett
- 1973
(Show Context)
Citation Context ...ne step further by recognizing that arbitrarily complex information processing could be carried out, in principle, by molecular machines of no greater complexity than those already observed in nature =-=[4, 5]-=-. Based on the intrinsic reversibility of chemical reactions, Bennett used this insight to give a thermodynamic argument that there is no fundamental energetic cost to computation — only a cost to era... |

469 |
Computation: finite and infinite machines
- Minsky
- 1967
(Show Context)
Citation Context ...re known to be less than Turing universal, 2 stacks are enough for universality. Similarly, while stack machines with just 1 symbol (also known as counter machines or register machines) are universal =-=[17]-=-, they are exponentially slower than Turing machines, and efficient simulation of134 L. Qian, D. Soloveichik, and E. Winfree Turing machines becomes possible only with 2 symbols or more. In fact, mul... |

353 |
Irreversibility and Heat Generation in the Computing Process
- Landauer
- 1961
(Show Context)
Citation Context ..., pp. 123–140, 2011. c○ Springer-Verlag Berlin Heidelberg 2011124 L. Qian, D. Soloveichik, and E. Winfree a cost to erase data. This conclusion derives from four principles: (1) as Landauer observed =-=[15]-=-, making a logically irreversible decision entails an energetic expenditure of kT ln 2, and thus there is an unavoidable cost to irreversible logical operations; (2) being logically reversible is not ... |

106 | Algorithmic self-assembly of DNA Sierpinski triangles
- ROTHEMUND, PAPADAKIS, et al.
(Show Context)
Citation Context ...uters implemented weaker models of computation such as digital circuits or finite state machines [22,3]. Two-dimensional molecular self-assembly is Turing universal [26], implementable with DNA tiles =-=[21]-=-, and can be physically and logically reversible [27], but it has the distinct disadvantage of storing the entire history of its computation within a supramolecular complex — it’s bulky. Recent work h... |

94 |
Time/space trade-offs for reversible computation
- Bennett
- 1989
(Show Context)
Citation Context ...th the mean speed being linear in the (arbitrarily near zero) energy expended per step; and (4) any logically irreversible computation can be recast with a minimal number of extra computational steps =-=[4, 6]-=- as a logically reversible computation that requires irreversible operations only when preparing input and output during repeated use. It’s intriguing to ask whether Landauer’s and Bennett’s principle... |

79 | A DNA and restriction enzyme implementation of Turing Machines. Unpublished manuscript
- Rothemund
- 1995
(Show Context)
Citation Context ...on of Turing machines. Early theoretical proposals made use of existing enzymes but required a series of laboratory manipulations to step the molecular Turing machines through their operational cycle =-=[20,2,23]-=-, while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures [14,4,11,28,12]. Experimental demonstra... |

77 | Rule-Based Modelling of Cellular Signalling
- Danos, Feret, et al.
(Show Context)
Citation Context ...s (e.g. endlocalized reactions, interior-localized reactions, polymer joining and scission) and the types of polymer structures (e.g. strictly linear, branched, networks with cycles) that are allowed =-=[13,8,11]-=-. All these natural CRN extensions can efficiently simulate Turing machines (c.f. [11]). However, whereas these extensions were designed to be general for modeling biochemical systems, our interest he... |

76 | The thermodynamics of computation: A review - Bennett - 1982 |

75 | On the Computational Power of DNA Annealing and Ligation - Winfree - 1996 |

69 |
Programmable and autonomous computing machine made of biomolecules
- Benenson, Paz-Elizur, et al.
- 2001
(Show Context)
Citation Context ...ymes or DNA nanostructures [14,4,11,28,12]. Experimental demonstrations of autonomous biomolecular computers implemented weaker models of computation such as digital circuits or finite state machines =-=[22,3]-=-. Two-dimensional molecular self-assembly is Turing universal [26], implementable with DNA tiles [21], and can be physically and logically reversible [27], but it has the distinct disadvantage of stor... |

69 | Simulations of Computing by Self-Assembly
- Winfree
(Show Context)
Citation Context ...s digital circuits or finite state machines [22,3]. Two-dimensional molecular self-assembly is Turing universal [26], implementable with DNA tiles [21], and can be physically and logically reversible =-=[27]-=-, but it has the distinct disadvantage of storing the entire history of its computation within a supramolecular complex — it’s bulky. Recent work has pointed to an alternative to geometrical organizat... |

66 |
Enzyme-free nucleic acid logic circuits
- Seelig, Soloveichik, et al.
- 2006
(Show Context)
Citation Context ...ymes or DNA nanostructures [14,4,11,28,12]. Experimental demonstrations of autonomous biomolecular computers implemented weaker models of computation such as digital circuits or finite state machines =-=[22,3]-=-. Two-dimensional molecular self-assembly is Turing universal [26], implementable with DNA tiles [21], and can be physically and logically reversible [27], but it has the distinct disadvantage of stor... |

53 | A universal molecular computer, in
- Beaver
- 1995
(Show Context)
Citation Context ...on of Turing machines. Early theoretical proposals made use of existing enzymes but required a series of laboratory manipulations to step the molecular Turing machines through their operational cycle =-=[20,2,23]-=-, while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures [14,4,11,28,12]. Experimental demonstra... |

51 |
BioNetGen: software for rule-based modeling of signal transduction based on the interactions of molecular domains
- Blinov, Faeder, et al.
- 2004
(Show Context)
Citation Context ... (e.g. end-localized reactions, interior-localized reactions, polymer joining and scission) and the types of polymer structures (e.g. strictly linear, branched, networks with cycles) that are allowed =-=[12, 7, 10]-=-. All these natural CRN extensions can efficiently simulate Turing machines (c.f. [10]). However, whereas these extensions were designed to be general for modeling biochemical systems, our interest he... |

39 |
Engineering entropy-driven reactions and networks catalyzed by DNA
- Zhang, Turberfield, et al.
(Show Context)
Citation Context ... occurs through domain + X. The top strand F2 will fall off when it’s only held to the bottom strand by the toehold and leave F1 as intermediate product I1. (This is the principle of toehold exchange =-=[29,30]-=-.) Compared to F1, the bottom strand of I1 has its first toehold covered and the second toehold revealed. Signal Y then binds to I1 at the second toehold, branch migrates to the 3 ′ end of + Y and kic... |

37 |
DNA as a universal substrate for chemical kinetics
- Soloveichik
- 2010
(Show Context)
Citation Context ...on networks (CRNs) with a finite number of species in a well-mixed solution are structurally simple enough (essentially geometry-free) that in principle arbitrary networks can be implemented with DNA =-=[25]-=-, yet they are (probabilistically) Turing universal [24]. This Turing universal computation using geometry-free chemical reaction networks is theoretically accurate and reasonably fast (only a polynom... |

30 |
Control of DNA strand displacement kinetics using toehold exchange
- Zhang, Winfree
(Show Context)
Citation Context ... occurs through domain + X. The top strand F2 will fall off when it’s only held to the bottom strand by the toehold and leave F1 as intermediate product I1. (This is the principle of toehold exchange =-=[29,30]-=-.) Compared to F1, the bottom strand of I1 has its first toehold covered and the second toehold revealed. Signal Y then binds to I1 at the second toehold, branch migrates to the 3 ′ end of + Y and kic... |

29 | A simple DNA gate motif for synthesizing large-scale circuits, DNA Computing
- Qian, Winfree
- 2009
(Show Context)
Citation Context ...l DNA species are of the same form, and allow the coupling of such formal reactions together. (They are also of the same form as signal species in another DNA strand displacement network architecture =-=[19]-=-, which allows even broader couplings.) All signal species have one short toehold domain in the middle, one long recognition domain “−”onthe5 ′ end and another long recognition domain “+” on the 3 ′ e... |

28 | DNA computers in vitro and vivo
- Smith, Schweitzer
- 1995
(Show Context)
Citation Context ...on of Turing machines. Early theoretical proposals made use of existing enzymes but required a series of laboratory manipulations to step the molecular Turing machines through their operational cycle =-=[20,2,23]-=-, while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures [14,4,11,28,12]. Experimental demonstra... |

22 | On the computational power of biochemistry
- Cardelli, Zavattaro
- 2008
(Show Context)
Citation Context ...ugh their operational cycle [19, 2, 21], while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures =-=[13, 3, 10, 26, 11]-=-. Two-dimensional molecular self-assembly is Turing universal [24], implementable with DNA tiles [20], and can be physically and logically reversible [25], but it has the distinct disadvantage of stor... |

22 | Computation with finite stochastic chemical reaction networks. Natural Computing: an international journal
- Soloveichik, Cook, et al.
- 2008
(Show Context)
Citation Context ...well-mixed solution are structurally simple enough (essentially geometry-free) that in principle arbitrary networks can be implemented with DNA [25], yet they are (probabilistically) Turing universal =-=[24]-=-. This Turing universal computation using geometry-free chemical reaction networks is theoretically accurate and reasonably fast (only a polynomial slowdown), but requires molecular counts (and theref... |

12 | Biological computing
- Kurtz, Mahaney, et al.
- 1997
(Show Context)
Citation Context ...rough their operational cycle [20,2,23], while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures =-=[14,4,11,28,12]-=-. Experimental demonstrations of autonomous biomolecular computers implemented weaker models of computation such as digital circuits or finite state machines [22,3]. Two-dimensional molecular self-ass... |

10 |
A 1-tape 2-symbol reversible Turing machine
- Morita, Shirasaki, et al.
- 1989
(Show Context)
Citation Context ... Turing machine with one tape that is bounded on the left and is infinite on the right; futher, we require that the Turing machine never tries to read past the left end of the tape. (For example ref. =-=[18]-=- describes such a Turing machine that is universal, although it is slow; multi-tape reversible Turing machines are faster, and can be similarly implemented with polymer CRNs.) We can consider three ty... |

9 |
A universal molecular computer. in DNA Based Computers
- Beaver
- 1995
(Show Context)
Citation Context ...on of Turing machines. Early theoretical proposals made use of existing enzymes but required a series of laboratory manipulations to step the molecular Turingmachines through their operational cycle =-=[19, 2, 21]-=-, while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures [13, 3, 10, 26, 11]. Two-dimensional mo... |

8 | Two-domain DNA strand displacement
- Cardelli
(Show Context)
Citation Context ... other reactions. In order to solve this problem, we develop a “history-free” implementation of arbitrary chemical reaction networks. (Cardelli has proposed an elegant and even further reduced scheme =-=[9]-=-, butit appears unsuitable for our polymer reaction construction.) We describe an irreversible scheme, that with slight modifications can become reversible. In addition to making the reversible schem... |

7 | Design of an autonomous DNA nanomechanical device capable of universal computation and universal translational motion
- Yin, Turberfield, et al.
- 2004
(Show Context)
Citation Context ...rough their operational cycle [20,2,23], while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures =-=[14,4,11,28,12]-=-. Experimental demonstrations of autonomous biomolecular computers implemented weaker models of computation such as digital circuits or finite state machines [22,3]. Two-dimensional molecular self-ass... |

6 | Turing complete catalytic particle computers
- Liekens, Fernando
- 2006
(Show Context)
Citation Context ...ction networks is theoretically accurate and reasonably fast (only a polynomial slowdown), but requires molecular counts (and therefore volumes) that grow exponentially with the amount of memory used =-=[16,24]-=-. In contrast, reaction networks using heterogeneous polymers — the simplest kind of geometrical organization, as in Bennett’s vision — can store all information as strings within a single polymer, th... |

3 | Towards programmable molecular machines
- Chen, De, et al.
- 2008
(Show Context)
Citation Context ...rough their operational cycle [20,2,23], while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures =-=[14,4,11,28,12]-=-. Experimental demonstrations of autonomous biomolecular computers implemented weaker models of computation such as digital circuits or finite state machines [22,3]. Two-dimensional molecular self-ass... |

2 |
algebras for DNA computing
- Cardelli
(Show Context)
Citation Context ...forward, the history-free signal strand motif simplifies the correspondence between the abstract CRN and the DNA implementation: each CRN species now corresponds to exactly one DNA species. Like ref. =-=[8]-=-, but unlike ref. [23], we use stochastic semantics in this paper where reactions manipulate integer molecular counts of the reacting species, rather than real-valued concentrations. The applicable ki... |

1 |
Molecular computing machines. Encyclopedia of Nanoscience and Nanotechnology
- Benenson, Shapiro
- 2004
(Show Context)
Citation Context ...ugh their operational cycle [19, 2, 21], while later theoretical proposals suggested how autonomous molecular Turing machines could be built but made use of hypothetical enzymes or DNA nanostructures =-=[13, 3, 10, 26, 11]-=-. Two-dimensional molecular self-assembly is Turing universal [24], implementable with DNA tiles [20], and can be physically and logically reversible [25], but it has the distinct disadvantage of stor... |

1 | E.: Molecular computing machines
- Benenson, Shapiro
- 2004
(Show Context)
Citation Context |