Lecture Notes in Computer Science 8553Commenced Publication in 1973Founding and Former Series Editors:Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen

Editorial Board

David HutchisonLancaster University, UK

Takeo KanadeCarnegie Mellon University, Pittsburgh, PA, USA

Josef KittlerUniversity of Surrey, Guildford, UK

Jon M. KleinbergCornell University, Ithaca, NY, USA

Alfred KobsaUniversity of California, Irvine, CA, USA

Friedemann MatternETH Zurich, Switzerland

John C. MitchellStanford University, CA, USA

Moni NaorWeizmann Institute of Science, Rehovot, Israel

Oscar NierstraszUniversity of Bern, Switzerland

C. Pandu RanganIndian Institute of Technology, Madras, India

Bernhard SteffenTU Dortmund University, Germany

Demetri TerzopoulosUniversity of California, Los Angeles, CA, USA

Doug TygarUniversity of California, Berkeley, CA, USA

Gerhard WeikumMax Planck Institute for Informatics, Saarbruecken, Germany

Oscar H. IbarraLila KariSteffen Kopecki (Eds.)

Unconventional Computationand Natural Computation

13th International Conference, UCNC 2014London, ON, Canada, July 14-18, 2014Proceedings

13

Volume Editors

Oscar H. IbarraUniversity of CaliforniaDepartment of Computer ScienceSanta Barbara, CA 93106, USAE-mail: ibarra@cs.ucsb.edu

Lila KariSteffen KopeckiUniversity of Western OntarioDepartment of Computer ScienceLondon, ON N6A 5B7, CanadaE-mail: {lila, steffen}@csd.uwo.ca

ISSN 0302-9743 e-ISSN 1611-3349ISBN 978-3-319-08122-9 e-ISBN 978-3-319-08123-6DOI 10.1007/978-3-319-08123-6Springer Cham Heidelberg New York Dordrecht London

Library of Congress Control Number: 2014940804

LNCS Sublibrary: SL 1 – Theoretical Computer Science and General Issues

© Springer International Publishing Switzerland 2014This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part ofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting, reproduction on microfilms or in any other physical way, and transmission or informationstorage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodologynow known or hereafter developed. Exempted from this legal reservation are brief excerpts in connectionwith reviews or scholarly analysis or material supplied specifically for the purpose of being entered andexecuted on a computer system, for exclusive use by the purchaser of the work. Duplication of this publicationor parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location,in ist current version, and permission for use must always be obtained from Springer. Permissions for usemay be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecutionunder the respective Copyright Law.The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoes not imply, even in the absence of a specific statement, that such names are exempt from the relevantprotective laws and regulations and therefore free for general use.While the advice and information in this book are believed to be true and accurate at the date of publication,neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors oromissions that may be made. The publisher makes no warranty, express or implied, with respect to thematerial contained herein.

Typesetting: Camera-ready by author, data conversion by Scientific Publishing Services, Chennai, India

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Preface

The International Conference on Unconventional Computation and Natural Com-putation, UCNC, is an interdisciplinary meeting where scientists with differentbackgrounds, yet sharing a common interest in novel forms of computation,human-designed computation inspired by nature, and the computational aspectsof processes taking place in nature, present their latest theoretical or experimen-tal results. The topics of the conference typically include:

Molecular computingQuantum computingOptical computingChaos computingPhysarum computingHyperbolic spacecomputation

Collision-basedcomputing

Super-Turingcomputation

Cellular automataNeural computationEvolutionarycomputation

Swarm intelligenceAnt algorithmsArtificial immunesystems

Artificial lifeMembrane computingAmorphous computing

Computationalsystems biology:

◦ genetic networks◦ protein–proteinnetworks

◦ transport networksComputationalneuroscience

Synthetic biologyCellular (in vivo)computing

The first edition of UCNC (formerly called Unconventional Models of Com-putation and Unconventional Computation) was held at the Centre for DiscreteMathematics and Theoretical Computer Science, Auckland, New Zealand, in1998, and the conference logo became the logo of its first host. Subsequent sitesof the conference were Brussels, Belgium, in 2000, Kobe, Japan, in 2002, Seville,Spain, in 2005, York, UK, in 2006, Kingston, Canada, in 2007, Vienna, Aus-tria, in 2008, Ponta Delgada, Portugal, in 2009, Tokyo, Japan, in 2010, Turku,Finland, in 2011, Orleans, France, in 2012, and Milan, Italy, in 2013.

The 13th edition in this conference series, UCNC 2014, was organized in Lon-don, Ontario, Canada, in the Deparment of Computer Science of the Universityof Western Ontario, during the week of July 14–18, 2014.

The meeting was pleased to have four distinguished invited speakers whopresented talks touching on several UCNC topics:

– Yaakov Benenson (ETH Zurich), “Molecular Computing Meets SyntheticBiology”

– Charles H. Bennett (IBM T. J. Watson Research Center), “From QuantumDynamics to Physical Complexity”

– Hod Lipson (Cornell University), “The Robotic Scientist: Distilling NaturalLaws from Experimental Data, from Cognitive Robotics to ComputationalBiology”

– Nadrian C. Seeman (New York University), “DNA: Not Merely the Secretof Life – Using the Information in DNA to Control Molecular Structure”

VI Preface

The meeting was also pleased to have three distinguished invited tutorialspeakers:

– Anne Condon (University of British Columbia), “Programming with Chem-ical Reaction Networks and DNA Strand Displacement Systems”

– Ming Li (University of Waterloo), “Approximating Semantics”– Tommaso Toffoli (Boston University), “Do We Compute to Live, or Live

to Compute? Entropy Pumps, Evolution vs Emergence, and the Risks ofSuccess”

This year, in response to the Call for Papers, there were 79 articles submit-ted by authors from 30 countries. Each paper was reviewed by at least threereferees and discussed by the members of the Program Committee. Finally, 31papers were selected for oral presentation at the conference and inclusion inthese proceedings.

The conference has a long history of hosting workshops. The 2014 edition inLondon hosted three workshops:

– “DNA Computing by Self-assembly,” organized by Matthew Patitz, withinvited speakers Scott Summers and Damien Woods (Tuesday, July 15)

– “Computational Neuroscience,” organized by Mark Daley, with invited speak-ers Randy McIntosh and William Cunningham (Thursday, July 17)

– “Unconventional Computation in Europe,” organized by Martyn Amos andSusan Stepney, with invited speaker Ricard Sole (Friday, July 18)

We are grateful for the support of the FIELDS Institute for Research inMathematical Sciences, the PERIMETER Institute for Theoretical Physics, theDepartment of Computer Science and the Faculty of Science of the Universityof Western Ontario, Research Western, IBM, and the Rotman Institute of Phi-losophy.

We thank all those who have contributed to this meeting. In particular, wethank the invited speakers, the contributing authors, the referees, the membersof the Program Committee, the members of the Steering Committee, the localorganizers and the Student Volunteer Team, all of whose efforts have contributedto the practical and scientific success of the meeting.

July 2014 Oscar H. IbarraLila Kari

Steffen Kopecki

Organization

Program Committee

Andrew Adamatzky University of the West of England, UKSelim G. Akl Queen’s University, CanadaEshel Ben-Jacob Tel Aviv University, IsraelCristian S. Calude University of Auckland, New ZealandJose Felix Costa IST University of Lisbon, PortugalErzsebet Csuhaj-Varju Eotvos Lorand University, HungaryAlberto Dennunzio University of Milano-Bicocca, ItalyMarco Dorigo Universite Libre de Bruxelles, BelgiumJerome Durand-Lose University of Orleans, FranceMasami Hagiya University of Tokyo, JapanOscar H. Ibarra University of California, Santa Barbara, USA

(Co-chair)Kazuo Iwama Kyoto University, JapanJarkko Kari University of Turku, FinlandLila Kari University of Western Ontario, Canada

(Co-chair)Viv Kendon University of Leeds, UKKamala Krithivasan IIT Madras, IndiaGiancarlo Mauri University of Milano-Bicocca, ItalyYongli Mi Hong Kong University of Science and

Technology, ChinaMario J. Perez-Jimenez University of Seville, SpainKai Salomaa Queen’s University, CanadaHava Siegelmann University of Massachusetts Amherst, USASusan Stepney University of York, UKDamien Woods California Institute of Technology, USAByoung-Tak Zhang Seoul National University, Korea

Steering Committee

Thomas Back Leiden University, The NetherlandsCristian S. Calude University of Auckland, New Zealand

(Founding Chair)Lov K. Grover Bell Labs, USANatasa Jonoska University of South Florida, USA (Co-chair)Jarkko Kari University of Turku, Finland (Co-chair)Lila Kari University of Western Ontario, Canada

VIII Organization

Seth Lloyd Massachusetts Institute of Technology, USAGiancarlo Mauri University of Milano-Bicocca, ItalyGheorghe Paun Institute of Mathematics of the Romanian

Academy, RomaniaGrzegorz Rozenberg Leiden University, The Netherlands

(Emeritus Chair)Arto Salomaa University of Turku, FinlandTommaso Toffoli Boston University, USACarme Torras Institute of Robotics and Industrial

Informatics, SpainJan van Leeuwen Utrecht University, The Netherlands

Organizing Committee

Mark Daley University of Western Ontario, CanadaHelmut Jurgensen University of Western Ontario, CanadaLila Kari University of Western Ontario, Canada (Chair)Steffen Kopecki University of Western Ontario, CanadaStephen Watt University of Western Ontario, Canada

Student Volunteer Team

Rallis Karamichalis University of Western Ontario, CanadaManasi Kulkarni University of Western Ontario, CanadaSrujan Kumar Enaganti University of Western Ontario, CanadaAmirhossein Simjour University of Western Ontario, CanadaTina Wu University of Western Ontario, Canada

Additional Reviewers

Arrighi, PabloBarr, KatieBazso, FulopBecker, FlorentBeretta, StefanoBesozzi, DanielaBienvenu, LaurentBorello, AlexBown, JamesCabajal, JuanCabessa, JeremieCapobianco, SilvioCastelli, Mauro

Dinneen, MichaelDomaratzki, MikeDoty, DavidElias, SusanEom, Hae-SungFates, NazimFazekas, Szilard ZsoltFerretti, ClaudioGajardo, AnahiGopinath, AshwinGorecki, JerzyGraca, DanielGuillon, Pierre

Gutierrez-Naranjo,Miguel A.

Han, Yo-SubHarju, TeroHickinbotham, SimonHirvensalo, MikaHunyadvari, LaszloImai, KatsunobuJonoska, NatasaKaced, TarikKanter, IdoKawamata, IbukiKo, Sang-Ki

Organization IX

Kobayashi, SatoshiKolonits, GaborKurka, PetrLakin, Matthew R.Lazar, Katalin A.Leporati, AlbertoLu, MingyangMa, XiongfengMacıas-Ramos,

Luis FelipeManea, FlorinManzoni, LucaMarchetti, LucaMarion, Jean-YvesMercas, RobertMetta, PadmaMeunier, Pierre-EtienneMiller, JulianMoisset De Espanes,

PabloMurphy, Niall

Mutyam, MadhuNair, AchuthsankarNeary, TurloughNobile, MarcoPalioudakis, AlexandrosPescini, DarioPolack, FionaPorreca, Antonio E.Poulding, SimonPoupet, VictorPradella, MatteoProctor, TimothyRahman, AfrozaRama, RaghavanRamanujan, AjeeshRichard, GaetanRiscos-Nunez, AgustınRomashchenko, AndreiRomero-Campero,

Francisco J.Salo, Ville

Saubion, FredericSchulman, RebeccaSeki, ShinnosukeSobot, RobertSong, BoshengStannett, MikeStefanovic, DarkoSummers, ScottSvozil, KarlSzabados, MichalThachuk, ChrisTichler, KrisztianTimmis, JonTimperley, ChrisUnold, OlgierdWinslow, AndrewYokomori, TakashiZandron, ClaudioZizza, Rosalba

X Organization

Sponsors

We deeply thank the sponsors that made UCNC 2014 possible.

Fields Institute for Research inMathematical Scienceshttp://www.fields.utoronto.ca/

Perimeter Institute forTheoretical Physicshttp://www.perimeterinstitute.ca/

IBMhttp://www.ibm.com/

University of Western Ontariohttp://www.uwo.ca/

Department of Computer ScienceFaculty of ScienceResearch WesternRotman Institute of Philosophy

Invited Talks

Molecular Computing Meets Synthetic Biology

Yaakov Benenson

Synthetic Biology GroupDepartment of Biosystems Science and Engineering

ETH Zurich, Mattenstrasse 26, 4058 Basel, Switzerland

Abstract. One of the motivations behind computing with molecules isto “computerize” living systems, for example to prevent disease or controlartificial tissues. Biology, however, is already very good at computing —the human brain being one example. Even on a single cell level informa-tion is constantly being processed, and the development of a functionalorganism from a single fertilized cell is controlled by an ingenious if onlypartially understood program encoded in DNA. Does this mean that theefforts to “write” new molecular programs are redundant? Not at all —natural programs have taken three billion years to evolve and, despitetheir beauty, are very difficult to alter in any way.

In my view the optimal approach is to balance the engineering prin-ciples inspired by computer science and engineering, such as universalmodels, reprogrammability, modularity, etc., with the harsh reality of celland organismal biology. The simple fact is that we do not know yet, evenat the theory level, whether it is possible to perform reliable informationprocessing in actual living cells as opposed to idealized “well-mixed re-actors”. Despite these limitations, the field of molecular computing incells, or biological computing, has made significant steps forward withnew design principles, new architectures, and new exciting experimentalresults. These developments also inform basic biological research.

From Quantum Dynamics to Physical

Complexity

Charles H. Bennett

IBM Research, Thomas J. Watson Research Center

Yorktown Heights, NY 10598, USA

Abstract. Quantum effects in information processing, aside from mak-ing possible feats like quantum cryptography and Shor’s factoring algo-rithm, have led to more coherent and powerful ways of thinking aboutinformation, computation, and cosmology. We review this approach, es-pecially the uniquely private form of correlation known as entanglement,whose very pervasiveness makes it hard to detect, allowing it to remainundiscovered until the 20th century. In combination with thermal dis-equilbrium, entanglement helps us understand why the future is moreuncertain than the past, and how our world produces structures thatare logically “deep”, having internal evidence of a complicated history,an idea that can be made precise using the tools of algorithmic informa-tion and computational complexity. Finally we consider the Boltzmannbrain problem afflicting many modern cosmologies, wheresimilar structures are predicted to fluctuate into existence even at ther-mal equilibrium, bearing false evidence of a complicated history thatnever happened.

The Robotic ScientistDistilling Natural Laws from Experimental Data, from

Cognitive Robotics to Computational Biology

Hod Lipson

Sibley School of Mechanical and Aerospace Engineering, Cornell University

Ithaca, NY 14853-7501, USA

Abstract. Can machines discover scientific laws automatically? Despitethe prevalence of computing power, the process of finding natural lawsand their corresponding equations has resisted automation. We will out-line a series of recent research projects, starting with self-reflecting roboticsystems, and ending with machines that can formulate hypotheses, de-sign experiments, and interpret the results, to discover new scientificlaws. We will then present examples from psychology to cosmology, fromclassical physics to modern physics, from big science to small science.

Reference

1. Schmidt, M., Lipson, H.: Distilling free-form natural laws from experimental data.Science 324(5923), 81–85 (2009)

DNA: Not Merely the Secret of LifeUsing the Information in DNA to Control Molecular

Structure�

Nadrian C. Seeman

Chemistry Department, New York University

New York, NY 10003, USA

Abstract. Our laboratory is investigating unusual DNA molecules inmodel systems that use synthetic molecules. A major effort in our labora-tory is devoted to DNA nanotechnology. The attachment of specific stickyends to a DNA branched junction enables the construction of stick fig-ures, whose edges are double-stranded DNA. This approach has alreadybeen used to assemble a cube, a truncated octahedron, nanomechanicaldevices and 2D crystals and 3D crystals from DNA. Ultimate goals forthis approach include the assembly of a biochip computer, nanoroboticsand nanofabrication and the exploitation of the rational synthesis of pe-riodic matter.

Thus, we build branched DNA species that can be joined usingWatson-Crick base pairing to produce N-connected objects and lattices. We haveused ligation to construct DNA topological targets, such as knots, poly-hedral catenanes, Borromean rings and a Solomon’s knot. Branched junc-tions with up to 12 arms have been made.

Nanorobotics is a key area of application. We have made robust 2-stateand 3-state sequence-dependent devices and bipedal walkers. We haveconstructed a molecular assembly line using a DNA origami layer andthree 2-state devices, so that there are eight different states representedby their arrangements. We have demonstrated that all eight productscan be built from this system.

A central goal of DNA nanotechnology is the self-assembly of periodicmatter. We have constructed 2D DNA arrays with designed patternsfrom many different motifs. We have used DNA scaffolding to organizeactive DNA components. We have used pairs of 2-state devices to capturea variety of different DNA targets.

One of the key aims of DNA-based materials research is to constructcomplex material patterns that can be reproduced. We have built sucha system from bent TX molecules, which can reach 2 generations ofreplication. This system represents a first step in self-reproducingmaterials. We are making progress towards selection of self-replicatingmaterials.

* This research has been supported by the NIGMS, NSF, ARO, ONR, DOE and theGordon and Betty Moore Foundation.

DNA: Not Merely the Secret of Life XVII

Recently, we have self-assembled a 3D crystalline array and havesolved its crystal structure to 3A resolution, using unbiased crystallo-graphic methods. We can use crystals with two molecules in the crys-tallographic repeat to control the color of the crystals. Thus, structuralDNA nanotechnology has fulfilled its initial goal of controlling the struc-ture of DNA in three dimensions. A new era in nanoscale control andmolecular programming awaits us.

Invited Tutorials

Programming with Chemical Reaction Networks

and DNA Strand Displacement Systems

Anne Condon

Department of Computer Science, University of British Columbia

Vancouver, BC, V6T 1Z4, Canada

Abstract. Chemical reaction networks (CRNs) and DNA strand dis-placement systems (DSDs) are widely-studied and useful models of molec-ular programming. In this tutorial, we introduce the models, illustratingthe expressive power of CRNs as a molecular programming languageand how CRNs can be compiled into lower-level, physically realizableDNA strand displacement systems. We characterize the power of CRNsin terms of well known complexity classes, describe connections withreversible and distributed computing models, and discuss limits to com-puting with CRNs. Finally, we discuss directions for future research thatcould advance our understanding of these models and the possibilitiesfor efficient molecular programs.

Bio-molecules do remarkable things in our cells, including information process-ing, communication and transportation. Recent technological advances have en-abled scientists to design and program simple DNA molecular systems witha variety of computational and functional capabilities, many of which alreadyexceed the roles of DNA in the cell. Bio-molecules are interesting to programbecause of their dynamic structural and material properties, because they en-able us to organize matter at the nano-scale and because they can naturallyinteract with biological systems at the cellular level. It is hard to imagine a fu-ture in which programming molecules will not be central to understanding andmediating cellular and other nano-scale processes.

So, how can we program molecules? Researchers in the field of DNA comput-ing and molecular programming have developed many creative approaches, alongwith experimental demonstrations of the viability of these approaches. In thistutorial I’ll focus on two such approaches, namely Chemical Reaction Networks(CRNs) [3, 9, 8] and DNA Strand Displacement Systems (DSDs) [12, 15].

CRNs are a distributed computing model in which, starting from an initialpool of molecules, consisting of duplicates from a finite set of species, reactionsensue that consume and produce species, thereby converging on an “outcome”pool that indicates the result of a computation. CRNs are interesting in partbecause they model chemical system kinetics - the basis for biological informationprocessing - and in part because they provide a very natural level of abstractionin which to design and reason about molecular processes.

DSD programs model a lower level of abstraction than CRNs. At their coreis a basic primitive whereby an initially unbound input DNA strand I binds to atemplate T , thereby displacing an output strand O that was initially bound to T

XXII A. Condon

so that O becomes unbound. Strand O can in turn act as input for a subsequentdisplacement. DSDs are collections of strands that can change configurations,i.e., which strands are bound and which are unbound, via successive strand dis-placements in a pre-programmed fashion, ultimately producing unbound strandsthat encode the result of a computation. There have been successful experimen-tal demonstrations of DSD designs that realize logic circuits and even artificialneurons [5, 6].

In this tutorial, we’ll describe why CRNs and DSDs are interesting program-ming models, what we know about effective ways to write CRN and DSD pro-grams, and what are important directions for further progress. For example,CRNs can be “compiled” into DSDs [1, 9, 10], CRNs and thus DSDs can in prin-ciple provide an energy efficient realization of CRNs [13, 7, 10, 14].

Several approaches for deterministicially computing with CRNs have beenstudied. In this context, resources such as time, volume (i.e., space needed tostore species as a computation proceeds), and energy are important. In somemodels, quantities are represented by the number of copies, or count, of amolecule. A computation is considered to have completed when the count ofdesignated output species is stable, i.e., will not change regardless of which ap-plicable reactions ensue. Such models are typically uniform, in the sense that thenumber of species needed to specify an algorithm is independent of the inputsize. Connections with population protocols, a distributed computing model, hasprovided valuable insights on the resources needed to compute with such models.Variants of these uniform models, in which molecular polymers can represent astack data structure, can simulate Turing-general models of computation [7, 4].In other, non-uniform models, the presence or absence of molecular species rep-resent bit values and thus, like (non-uniform) circuit models of computation, thenumber of species needed for a computation is a function of the input length.It is possible to design non-uniform DSDs that “recycle” molecules by runningreversible reactions or displacements in both forwards and reverse directions,so that t steps of the system use just O(log t) molecules [2, 11] and thus havelimited volume.

A limitation of some designed DSDs [4, 2] is that, in order for them to com-pute correctly, a single copy of some reactants should be present initially. It iscurrently impractical to obtain the exact numbers in a wet lab. When multiplecopies of all initial molecules are present, correctness requires that the lengthof the shortest sequence of reactions needed to produce any given molecule isbounded by a polynomial function of the (appropriately measured) size of theCRN [2].

There are many interesting directions for future research. Techniques areneeded for establishing the correctness of even quite simple CRNs. Better mech-anisms are needed for translating CRNs to physically realizable DSDs that canbe implemented robustly in the face of errors. More work is needed to under-stand what can be computed reversibly and with limited volume. Such research,grounded in an appreciation for and understanding of thermodynamics and

Programming with Chemical Reaction Networks XXIII

kinetics, as well as the potential and limitations of experimental systems, canhelp lay the foundations of an exciting new discipline.

References

1. Chen, H.-L., Doty, D., Soloveichik, D.: Deterministic function computation withchemical reaction networks. In: Stefanovic, D., Turberfield, A. (eds.) DNA 18 2012.LNCS, vol. 7433, pp. 25–42. Springer, Heidelberg (2012)

2. Condon, A., Hu, A.J., Manuch, J., Thachuk, C.: Less haste, less waste: On recyclingand its limits in strand displacement systems. J. R. Soc. Interface (2012)

3. Cook, M., Soloveichik, D., Winfree, E., Bruck, J.: Programmability of chemicalreaction networks. Algorithmic Bioprocesses 133, 543–584 (2009)

4. Qian, L., Soloveichik, D., Winfree, E.: Efficient Turing-universal computation withDNA polymers. In: Sakakibara, Y., Mi, Y. (eds.) DNA 16 2010. LNCS, vol. 6518,pp. 123–140. Springer, Heidelberg (2011)

5. Qian, L., Winfree, E.: Scaling up digital circuit computation with DNA stranddisplacement cascades. Science 332, 1196–1201 (2011)

6. Qian, L., Winfree, E., Bruck, J.: Neural network computation with DNA stranddisplacement cascades. Nature 475, 368–372 (2011)

7. Seelig, G., Soloveichik, D., Zhang, D.Y., Winfree, E.: Enzyme-free nucleic acid logiccircuits. Science 314(5805), 1585–1588 (2006)

8. Soloveichik, D.: Robust stochastic chemical reaction networks and bounded tau-leaping. J. Comput. Biol. 16(3), 501–522 (2009)

9. Soloveichik, D., Cook, M., Winfree, E., Bruck, J.: Computation with finite stochasticchemical reaction networks. Nat. Comp. 7, 615–633 (2008)

10. Soloveichik, D., Seelig, G., Winfree, E.: DNA as a universal substrate for chemicalkinetics. Proc. Nat. Acad. Sci. USA 107(12), 5393–5398 (2010)

11. Thachuk, C., Condon, A.: Space and energy efficient computation with DNA stranddisplacement systems. In: Stefanovic, D., Turberfield, A. (eds.) DNA 2012. LNCS,vol. 7433, pp. 135–149. Springer, Heidelberg (2012)

12. Yurke, B., Mills, A.P.: Using DNA to power nanostructures. Genet. Program Evolv-able Mach. 4(2), 111–122 (2003)

13. Yurke, B., Turberfield, A.J., Mills, A.P., Simmel, F.C., Neumann, J.L.: A DNA-fuelled molecular machine made of DNA. Nature 406, 605–608 (2000)

14. Zhang, D.Y., Seelig, G.: Dynamic DNA nanotechnology using strand displacementreactions. Nature Chemistry 3, 103–113 (2011)

15. Zhang, D.Y., Turberfield, A.J., Yurke, B., Winfree, E.: Engineering entropy-drivenreactions and networks catalyzed by DNA. Science 318, 1121–1125 (2007)

Approximating Semantics�

Ming Li

David R. Cheriton School of Computer Science, University of Waterloo

Waterloo, Ontario N2L 3G1, Canada

Abstract. Latent search engines and question-answering (QA) enginesfundamentally depend on our intuitive notion of semantics and semanticdistance. However, such a semantic distance is likely undefinable, cer-tainly un-computable, and often blindly approximated. Can we developa theoretical framework for this area?

We will describe a theory, using the well-defined information distance,to approximate the elusive semantic distance such that it is mathemat-ically proven that our approximation is “better than” any computableapproximation of the intuitive concept of semantic distance. Althoughinformation distance itself is obviously also not computable, it does allowa natural approximation by compression, especially with the availabilityof big data. We will then describe a natural language encoding systemto implement our theory followed by experiments on a QA system.

* This work is supported in part by NSERC Grant OGP0046506, OCRiT Grant115354, IDRC Research Chair in Information Technology, Project Number:104519-006, CFI ORF equipment grant, and the CRC Program.

Do We Compute to Live, or Live to Compute?Entropy Pumps, Evolution vs Emergence,

and the Risks of Success

Tommaso Toffoli

Department of Electrical and Computer Engineering

Boston University, Boston, MA 02215, USA

Abstract. We shall show that, in a “multiuser” world, strict immortal-ity in the long term is a contradiction in terms. Next best, then, what isa good strategy for at least some part of me to still be present in a largeproportion of samples of the future? Perhaps long individual life, manyidentical clones, continual repair, sexual reproduction, uploading myselfto the cloud, or what else?

Even if for sake of argument I grant that “survival of the fittest”is a mere tautology, so that ‘fittest’ just means “whoever survives” (afatalistic que sera sera; cf. analogous tautological constructs such as “theinvisible hand of the marketplace”), I am still left with the fundamentalquestion: What kinds of structure have what it takes to survive in mycurrent environment? In other words, for a given natural or artificialenvironment, what properties of a complex structure give it a chance toenjoy permanence in it? It is remarkable, but perhaps not too surprising,that this problem may have quite different solutions depending on thetime scale one has in mind (as we shall see, there are “greedy” strategiesthat can promise short-term permanence but virtually guarantee long-term disappearance).

‘Apparition’ and ‘permanence’ are key features of all sorts of emergentsystems — and these are found virtually whenever there is available anentropy pump. Lifelike systems are emergent systems that have beencaught in a special kind of positive-feedback loop: a runaway (at leastfor a while) loop with branching tracks, so that from the same initialconditions different “historical developments” are potentially available.

Evolution may be seen as a special case of emergence, namely, thedevelopment and interplay of a tangled hierarchy of emergent systemssome of which are lifelike. We shall be specially interested in the natureof the entropy pumps on which emergent systems are dependent, and inthe hierarchy of entropy pumps — the “entropy cascade” — that drivesevolution. In this context, we shall present a novel way to look at bothentropy and computation.

Table of Contents

Five Nodes Are Sufficient for Hybrid Networks of EvolutionaryProcessors to Be Computationally Complete . . . . . . . . . . . . . . . . . . . . . . . . . 1

Artiom Alhazov, Rudolf Freund, and Yurii Rogozhin

Learning Two-Input Linear and Nonlinear Analog Functions with aSimple Chemical System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

Peter Banda and Christof Teuscher

Scaled Tree Fractals Do not Strictly Self-assemble . . . . . . . . . . . . . . . . . . . . 27Kimberly Barth, David Furcy, Scott M. Summers, and Paul Totzke

GUBS a Language for Synthetic Biology: Specification andCompilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Adrien Basso-Blandin and Franck Delaplace

Modeling Syntactic Complexity with P Systems: A Preview . . . . . . . . . . . 54Gemma Bel Enguix and Benedek Nagy

Simulating Cancer Growth Using Cellular Automata to DetectCombination Drug Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Jenna Butler, Frances Mackay, Colin Denniston, and Mark Daley

Towards an MP Model for B Lymphocytes Maturation . . . . . . . . . . . . . . . 80Alberto Castellini, Giuditta Franco, Vincenzo Manca,Riccardo Ortolani, and Antonio Vella

Pseudo-inversion on Formal Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93Da-Jung Cho, Yo-Sub Han, Shin-Dong Kang, Hwee Kim,Sang-Ki Ko, and Kai Salomaa

Reverse-Engineering Nonlinear Analog Circuits with EvolutionaryComputation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Theodore W. Cornforth and Hod Lipson

Steps toward Developing an Artificial Cell Signaling Model Applied toDistributed Fault Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Dipankar Dasgupta and Guilherme Costa Silva

Dynamic Adaptive Neural Network Array . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Mark E. Dean, Catherine D. Schuman, and J. Douglas Birdwell

Producibility in Hierarchical Self-assembly . . . . . . . . . . . . . . . . . . . . . . . . . . 142David Doty

Unconventional Arithmetic: A System for Computation Using ActionPotentials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

Jonathan Edwards, Simon O’Keefe, and William D. Henderson

XXVIII Table of Contents

Reservoir Computing Approach to Robust Computation UsingUnreliable Nanoscale Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Alireza Goudarzi, Matthew R. Lakin, and Darko Stefanovic

On DNA-Based Gellular Automata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177Masami Hagiya, Shaoyu Wang, Ibuki Kawamata, Satoshi Murata,Teijiro Isokawa, Ferdinand Peper, and Katsunobu Imai

Doubles and Negatives are Positive (in Self-assembly) . . . . . . . . . . . . . . . . 190Jacob Hendricks, Matthew J. Patitz, and Trent A. Rogers

On String Languages Generated by Sequential Spiking Neural PSystems Based on Maximum Spike Number . . . . . . . . . . . . . . . . . . . . . . . . . 203

Keqin Jiang, Yuzhou Zhang, and Linqiang Pan

Languages Associated with Crystallographic Symmetry . . . . . . . . . . . . . . . 216Natasa Jonoska, Mile Krajcevski, and Gregory McColm

An Energy-Efficient Computing Approach by Filling the ConnectomeGap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229

Yasunao Katayama, Toshiyuki Yamane, and Daiju Nakano

Fast Arithmetic in Algorithmic Self-assembly . . . . . . . . . . . . . . . . . . . . . . . . 242Alexandra Keenan, Robert Schweller, Michael Sherman, andXingsi Zhong

Pattern Formation by Spatially Organized Approximate MajorityReactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254

Matthew R. Lakin and Darko Stefanovic

Mecobo: A Hardware and Software Platform for In MaterioEvolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267

Odd Rune Lykkebø, Simon Harding, Gunnar Tufte, andJulian F. Miller

Compact Realization of Reversible Turing Machines by 2-StateReversible Logic Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280

Kenichi Morita and Rei Suyama

Universal Computation in the Prisoner’s Dilemma Game . . . . . . . . . . . . . . 293Brian Nakayama and David Bahr

Exact Simulation of One-Dimensional Chaotic Dynamical SystemsUsing Algebraic Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305

Asaki Saito and Shunji Ito

Artificial Astrocyte Networks, as Components in Artificial NeuralNetworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316

Zahra Sajedinia

Table of Contents XXIX

Quantum, Stochastic, and Pseudo Stochastic Languages with FewStates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327

Arseny M. Shur and Abuzer Yakaryılmaz

Phase Transition and Strong Predictability . . . . . . . . . . . . . . . . . . . . . . . . . . 340Kohtaro Tadaki

PHLOGON: PHase-based LOGic using Oscillatory Nano-systems . . . . . . 353Tianshi Wang and Jaijeet Roychowdhury

Size-Separable Tile Self-assembly: A Tight Bound for Temperature-1Mismatch-Free Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

Andrew Winslow

Development of Physical Super-Turing Analog Hardware . . . . . . . . . . . . . . 379A. Steven Younger, Emmett Redd, and Hava Siegelmann

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393