Artificial Intelligence in Neuroscience and Systems Biology

120 Pages·2010·7.37 MB·English

Advances in Artificial Intelligence
Artificial Intelligence in
Neuroscience and Systems Biology:
Lessons Learnt, Open Problems,
and the Road Ahead
Guest Editors: Daniel Berrar, Naoyuki Sato, and Alfons Schuster
Artificial Intelligence in Neuroscience
and Systems Biology: Lessons Learnt,
Open Problems, and the Road Ahead
Advances in Artificial Intelligence
Artificial Intelligence in Neuroscience
and Systems Biology: Lessons Learnt,
Open Problems, and the Road Ahead
Guest Editors: Daniel Berrar, Naoyuki Sato,
and Alfons Schuster
Copyright©2010HindawiPublishingCorporation.Allrightsreserved.
Thisisaspecialissuepublishedinvolume2010of“AdvancesinArtificialIntelligence.”Allarticlesareopenaccessarticlesdistributed
undertheCreativeCommonsAttributionLicense,whichpermitsunrestricteduse,distribution,andreproductioninanymedium,
providedtheoriginalworkisproperlycited.
Advances in Artificial Intelligence
Editorial Board
MohamedM.Alimi,Tunisia RattikornHewett,USA JeffZ.Pan,UK
EduardoAlonso,UK PascalHitzler,USA DavePatterson,UK
AladdinAyesh,UK JunHong,UK MartinPelikan,USA
BikramjitBanerjee,USA FakhreddineKarray,Canada GuilinQi,Germany
RomanBartak,CzechRepublic WeiruLiu,UK AlfonsSchuster,UK
DanielBerrar,UK BruceJ.MacLennan,USA AlaaFathySheta,Jordan
CyrilleBertelle,France MarkMcCartney,UK ShiliangSun,China
DjamelBouchaffra,USA GerardMcKee,UK PeterTino,UK
Anto´nioD.P.Correia,Portugal GiorgioMetta,Italy VincentVidal,France
D.Girimonte,TheNetherlands IanMitchell,UK JonathanVincent,UK
DavidGlass,UK RichardMitchell,UK FaroukYalaoui,France
BernhardGraimann,Germany IvetaMrazova,CzechRepublic FilipZelezny,CzechRepublic
ChristianHo¨lscher,UK DebajyotiMukhopadhyay,India
JunHe,UK BarryO’Sullivan,Ireland
Contents
ArtificialIntelligenceinNeuroscienceandSystemsBiology:LessonsLearnt,OpenProblems,andthe
RoadAhead,DanielBerrar,NaoyukiSato,andAlfonsSchuster
Volume2010,ArticleID578309,2pages
QuoVadis,ArtificialIntelligence?,DanielBerrar,NaoyukiSato,andAlfonsSchuster
Volume2010,ArticleID629869,12pages
WhereArtificialIntelligenceandNeuroscienceMeet:TheSearchforGroundedArchitecturesof
Cognition,rankvanderVelde
Volume2010,ArticleID918062,18pages
RecurrenceQuantificationAnalysisofSpontaneousElectrophysiologicalActivityduringDevelopment:
CharacterizationofInVitroNeuronalNetworksCulturedonMultiElectrodeArrayChips,
AntonioNovellinoandJose´-ManuelZald´ıvar
Volume2010,ArticleID209254,10pages
SimulationofHumanEpisodicMemorybyUsingaComputationalModeloftheHippocampus,
NaoyukiSatoandYokoYamaguchi
Volume2010,ArticleID392868,10pages
ApplicationofGameTheorytoNeuronalNetworks,AlfonsSchusterandYokoYamaguchi
Volume2010,ArticleID521606,12pages
ConstraintsofBiologicalNeuralNetworksandTheirConsiderationinAIApplications,RichardStafford
Volume2010,ArticleID845723,6pages
InvestigatingtheUnderlyingIntelligenceMechanismsoftheBiologicalOlfactorySystem,
YoshinariMakinoandMasafumiYano
Volume2010,ArticleID478107,9pages
BootstrapLearningandVisualProcessingManagementonMobileRobots,MohanSridharan
Volume2010,ArticleID765876,20pages
FromExperimentalApproachestoComputationalTechniques:AReviewonthePredictionof
Protein-ProteinInteractions,FionaBrowne,HuiruZheng,HaiyingWang,andFranciscoAzuaje
Volume2010,ArticleID924529,15pages
HindawiPublishingCorporation
AdvancesinArtificialIntelligence
Volume2010,ArticleID578309,2pages
doi:10.1155/2010/578309
Editorial
Artificial Intelligence in Neuroscience and Systems Biology:
Lessons Learnt, Open Problems, and the Road Ahead
DanielBerrar,1,2NaoyukiSato,3andAlfonsSchuster4,5
1SystemsBiologyResearchGroup,CentreforMolecularBiosciences,SchoolofBiomedicalSciences,UniversityofUlster,CromoreRoad,
BT521SA,Coleraine,NorthernIreland
2SystemsBiologyDepartment,CancerInstitute,JapaneseFoundationforCancerResearch,Tokyo,Japan
3DepartmentofComplexSystems,FutureUniversityHakodate,116-2Kamedanakano-cho,Hakodate,Hokkaido041-8655,Japan
4SchoolofComputingandMathematics,FacultyofComputingandEngineering,UniversityofUlster,ShoreRoad,New-Townabbey,
Co.Antrim,BT370QB,NorthernIreland
5LaboratoryforDynamicsofEmergentIntelligence,RIKENBrainScienceInstitute,Wako-shi,Saitama351-0198,Japan
CorrespondenceshouldbeaddressedtoDanielBerrar,[email protected]
Received31January2010;Accepted31January2010
Copyright©2010DanielBerraretal.ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense,
whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
“Wecanonlyseeashortdistanceahead,butwe multitudeofviewsandcontributionsfromawidespectrum
canseeplentytherethatneedstobedone.” of contributors. This special issue, therefore, aims to create
AlanM.Turing
a centre of gravity pulling together academic researchers
Since its conception in the mid 1950s, artificial intelligence andindustrypractitionersfromavarietyofareasandback-
withitsgreatambitiontounderstandintelligence,itsorigin, groundstoshareresultsofcurrentresearchanddevelopment
and creation, in natural and artificial environments alike, andtodiscussexistingandemergingtheoreticalandpractical
hasbeenatrulymultidisciplinaryfieldthatreachesoutand problemsinartificialintelligence,neuroscience,andsystems
is inspired by a great diversity of other fields in perpetual biologytransportingthembeyondtheeventhorizonoftheir
motion. Rapid advances in research and technology in individualdomains.
variousfieldshavecreatedenvironmentsintowhichartificial Ifthecontributionsinthisspecialissuearetobeclassified
intelligence could embed itself naturally and comfortably. (crudely) according to their main thrust, then the articles
Neurosciencewithitsdesiretounderstandnervoussystems close to the neuroscience camp are devoted to the themes
ofbiologicalorganismsandsystembiologywithitslonging of (i) the characterization of in vitro neuronal networks
tocomprehend,holistically,themultitudeofcomplexinter- culturedonmultielectrodearray(MEA)chips,(ii)thesimu-
actionsinbiologicalsystemsaretwosuchfields.Theytarget lationofhumanepisodicmemorybyusingacomputational
idealsartificialintelligencehasdreamtaboutforalongtime modelofthehippocampus,and(iii)informationprocessing
including the computer simulation of an entire biological in natural and artificial olfactory systems. Novellino and
brain or the creation of new lifeforms from manipulations Zald´ıvarproposeacombinationofrecurrencequantification
oncellularandgeneticinformationinthelaboratory. analysisbasedonrecurrenceplotsandconventionalstatisti-
The scope for artificial intelligence, neuroscience, and calanalysisforneuronalelectrophysiology.Theyinvestigate
computational systems biology is extremely wide. The their approach by studying the variation of spontaneous
motivation of this special issue is to create a bird-eye view electrophysiological activity of in vitro neuronal networks
on areas and challenges where these fields overlap in their on multielectrode array chips. Sato and Yamaguchi review
definingambitionsandwherethesefieldsmaybenefitfrom computationalmodelsofthehippocampusanddiscusstheir
asynergeticmutualexchangeofideas.Therationalebehind owncomputationalmodelofhumanepisodicmemorybased
this special issue is that a multidisciplinary approach in onneuralsynchronization.Usingcomputersimulationsand
modern artificial intelligence, neuroscience, and systems human eye movement data, they demonstrate the validity
biologyisessentialandthatprogressinthesefieldsrequiresa of their model to predict human memory recall. From
2 AdvancesinArtificialIntelligence
an evolutionary perspective, Stafford reviews the biological
constraints of the physical properties of neurons and the
implicationfortheconstructionofartificialneuralnetworks.
van der Velde discusses fundamentals of human cognitive
processes and proposes models for grounded architectures
of cognition. Makino and Yano investigate olfaction as a
relatively simple biological information processing system
and report their computational works with a focus on the
temporaldimension.
The article by Browne et al. reviews computational
techniques to infer protein-protein interaction networks,
whichmayhelpdeciphernoveldrugtargets.
Two articles are supported by a strong background
in artificial intelligence and focus on learning algorithms.
SchusterandYamaguchiinvestigategametheoreticconcepts
and present a novel learning algorithm for a paired neuron
system. Sridharan reports on novel bootstrapped learning
techniquestoprocessvisualinputsthatallowamobilerobot
toautonomouslyadaptitsbehaviortoilluminationchanges.
In summary, this special issue informs the research
community at large about an exciting and stimulating
relationshipbetweenartificialintelligence,neuroscience,and
systems biology. The special issues provides access to many
state-of-the-art theoretical and applied problems in these
hugelyexcitingfieldsthataresorelevantformodernscience.
This special issue is also intended as a platform to bridge
cultural and technological gaps between these disciplines.
Ultimately, the contributions in this special issue should
convey to its readership the enthusiasm the editors and
authorsofthisissuesharefortheirrespectivefields.
DanielBerrar
NaoyukiSato
AlfonsSchuster
HindawiPublishingCorporation
AdvancesinArtificialIntelligence
Volume2010,ArticleID629869,12pages
doi:10.1155/2010/629869
Research Article
Quo Vadis, Artificial Intelligence?
DanielBerrar,1,2NaoyukiSato,3andAlfonsSchuster4,5
1SystemsBiologyResearchGroup,CentreforMolecularBiosciences,SchoolofBiomedicalSciences,UniversityofUlster,
CromoreRoad,BT521SAColeraine,UK
2SystemsBiologyDepartment,CancerInstitute,JapaneseFoundationforCancerResearch,Tokyo1358550,Japan
3DepartmentofComplexSystems,FutureUniversityHakodate,116-2Kamedanakano-cho,Hakodate,Hokkaido041-8655,Japan
4SchoolofComputingandMathematics,FacultyofComputingandEngineering,UniversityofUlster,ShoreRoad,
Newtownabbey,CountyAntrimBT370QB,UK
5LaboratoryforDynamicsofEmergentIntelligence,RIKENBrainScienceInstitute,Wako-shi,Saitama351-0198,Japan
CorrespondenceshouldbeaddressedtoDanielBerrar,[email protected]
Received9October2009;Accepted4January2010
AcademicEditor:DavidGlass
Copyright©2010DanielBerraretal.ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense,
whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
Since its conception in the mid 1950s, artificial intelligence with its great ambition to understand and emulate intelligence in
naturalandartificialenvironmentsalikeisnowatrulymultidisciplinaryfieldthatreachesoutandisinspiredbyagreatdiversityof
otherfields.Rapidadvancesinresearchandtechnologyinvariousfieldshavecreatedenvironmentsintowhichartificialintelligence
couldembeditselfnaturallyandcomfortably.Neurosciencewithitsdesiretounderstandnervoussystemsofbiologicalorganisms
andsystemsbiologywithitslongingtocomprehend,holistically,themultitudeofcomplexinteractionsinbiologicalsystemsare
twosuchfields.Theytargetidealsartificialintelligencehasdreamtaboutforalongtimeincludingthecomputersimulationofan
entirebiologicalbrainorthecreationofnewlifeformsfrommanipulationsofcellularandgeneticinformationinthelaboratory.
Thescopeforartificialintelligenceinneuroscienceandsystemsbiologyisextremelywide.Thisarticleinvestigatesthestanding
ofartificialintelligenceinrelationtoneuroscienceandsystemsbiologyandprovidesanoutlookatnewandexcitingchallenges
forartificialintelligenceinthesefields.Thesechallengesinclude,butarenotnecessarilylimitedto,theabilitytolearnfromother
projectsandtobeinventive,tounderstandthepotentialandexploitnovelcomputingparadigmsandenvironments,tospecifyand
adheretostringentstandardsandrobuststatisticalframeworks,tobeintegrative,andtoembraceopennessprinciples.
1.Introduction animalsareintelligenttosomedegree,byintelligencepeople
often assume the same scope of intelligence as can be seen
Sinceitsfoundation,whichisoftenassociatedwithaconfer- in general purpose human action (e.g., [4] mentions the
enceheldatDartmouthCollegeinNewHampshirein1956, creation of artifacts that are capable of mimicking and
artificial intelligence (AI) can look back on a rather excit- expressinghumanintelligence,thought,consciousness,and
ing and successful (though not unproblematic) career.Per emotion). Although the field has been exposed to some
definition, artificial intelligence—which may be described (often well-deserved) criticisms over the years, it may be
as a field of study embracing concepts, methodologies, and fair to say that AI is a truly great scientific endeavor
techniques as part of a computer program that exhibits that reaches out and embraces a great variety of scientific
characteristicsakintointelligentbehavior[1]—isasensitive disciplines ranging from areas that are mathematically and
and vulnerable field as the term intelligence itself lacks conceptually well defined (e.g., computer science, machine
a generally (ideally universally) acknowledged definition. learning, biology, and neuroscience) to areas that are more
There is, of course, no shortage of characterizations for difficulttoquantifyanddealwith(e.g.,philosophyofmind,
thetermintelligence(e.g.,[2,pages6–21]or[3,pages1–4]), cognitive science, and emotional intelligence). Of course, it
at large; however, the term remains a complicated, subtle, is not possible here to acknowledge adequately the many
and malleable topic. In addition, although, arguably, many contributions(includingthosecomingfromrapidadvances
2 AdvancesinArtificialIntelligence
in computer hardware and software) that shaped AI into to emphasize that these areas include mainstream utilities
the exciting discipline it is today, but it may be possible to such as the World Wide Web (e.g., the so-called intelligent
categorize, crudely, the development of AI into a few key web draws its power from algorithms that process infor-
stages(e.g.,see[3,pages1–24]). mation intelligently—Google’s page ranking algorithm or
Early AI projects had to come to terms with the many algorithms discovering matches on social-networking sites
challenges related to the production of knowledge-based are two examples [6]), toys and gadgets (e.g., computer
(expert) systems, including the acquisition of knowledge, games [7] or the Lego Mindstorms Robotics Invention
its representation and evolution, among other challenges. System [8]), extremely helpful human-computer inter-
The General Problem Solver, the early DENDRAL and faces (e.g., user-friendly interfaces using natural language
MYCIN expert systems, or the ELIZA program may stand recognition and visualizations including brain-computer
representativelyforthisperiod.TheGeneralProblemSolver interfaces [9, 10]), or humanoid robots functioning in
is a general purpose program to simulate human decision- various roles as partners for people in the immediate
making, thinking, and reasoning; DENDRAL (a system for environment of human beings (e.g., AIBO, ASIMO) (e.g.,
analyzingchemicals)andMYCIN(asystemforthediagnosis http://world.honda.com/ASIMO/).Artificialintelligencehas
of infectious blood diseases) are knowledge-based expert even reached out beyond earthly confines and is heav-
systemsandfocusontheproblemsolvinginspecificdomains ily utilized for numerous tasks by several space agencies
rather than on a general problem-solving strategy that is includingtheEuropeanSpaceAgency(ESA)(e.g.,[11])and
applicabletomany(all)domains;ELIZAisaprogramrelated NASA (National Aeronautics and Space Administration)
to natural language processing (and the Turing test) where (e.g.,http://www-aig.jpl.nasa.gov/).
humans interacting with a machine in a question-answer The previous points may suggest that contemporary AI
script-basedmodewereledtobelievethattheywereactually should find itself in a comfortable situation from which it
participatinginahuman-humaninteraction. should firmly and enthusiastically continue its cause—the
Artificial intelligence received a major boost in the mid study and creation of artificial entities that are capable of
1980s from works on artificial neural networks (ANNs). It expressinghumanintelligence,thought,consciousness,and
ispossibletosaythatsincetheearlydaysoftheperceptron emotion. The paper argues that this may not necessarily be
the story of artificial neural networks is one of the most so and that AI may need to be cautious and perhaps alert
successful chapters in the voluminous book of AI. Since to some degree about its standing because just around the
the emergence of several techniques, many of them united corner there are extremely exciting and powerful fields that
underthesoftcomputingparadigmthenprovidedAIwitha touchuponareas—andbegintooutshineAIinareas—that
capableandflexiblerepositoryforboththeoreticalresearch are at the very heart of AI itself. These fields include, but
as well as hands-on problem-solving applications. Possibly, are not necessarily limited to, neuroscience [12], synthetic
fromanapplicationpointofview,thiswasoneofthemost biology,andsystemsbiology[13].Thesedisciplinescomprise
exciting times for AI and the term knowledge engineering, several subfields with boundaries that are often blurred.
which suggests that systems showing some degree of intel- They also exemplify many modern research endeavors that
ligent problem-solving ability could be assembled from a are characterized by their complexity and a truly interdis-
toolboxofavailabletechniques,maycapturethisexcitement ciplinary nature that draws upon the expertise of scientists
quitewell. fromawiderangeofacademicbackgrounds.
In more recent times, AI has made it into the limelight The forthcoming sections investigate how AI is situated
through the DARPA (Defense Advanced Research Projects in this extended environment. Initially, Section2 takes a
Agency) Grand Challenges (http://www.darpa.mil/grand closer look at the interplay between AI, neuroscience,
challenge/index.asp). synthetic biology, and systems biology. Section3 identifies
In these challenges, vehicles had to navigate auton- several challenging hurdles in this interplay and includes
omously in increasingly challenging environments in an suggestionsforhowthesehurdlesmaybeovercomeinorder
intelligent manner, avoiding obstacles and solving prob- to create a synergetic and productive environment that is
lems of increasing levels of complexity, without human beneficial and supportive to practitioners working in these
intervention (e.g., [5]). Without hesitation one needs to fields. Section4 provides concluding remarks and ends the
acknowledge the achievements in these tasks, paralleled paperwithasummary.
perhapsintheirprestigeandpopularitybytheoutstanding
successes of IBM’s Deep Blue supercomputer (the machine
that recorded the first win in a game of chess against a 2.QuoVadis,ArtificialIntelligence?
reigningWorldChessChampion,GarryKasparov,in1996)
and the Deep Fritz chess engine (the commercial chess A starting point for an answer to this question may be
engine that triumphed 4-2 in December 2006 against the one of the most successful scientific endeavors in the last
reigningchampionVladimirKramnikinasix-gamematch) century—theHumanGenomeProject.TheHumanGenome
(http://www.research.ibm.com/deepblue/). It is beyond the Project achieved its goal, the definition of the sequence of
scope of this paper to pay credit to the great variety chemical base pairs which make up DNA, on 26 June 2000
of important and popular areas in which AI has made with the publication of a first draft of the human genome
substantialcontributions(oftenwithoutbeingdulycredited, (acknowledged in special issues in Nature [14] and Science
taken for granted, or simply neglected) but it is helpful [15]).Simplyspeaking,thehumangenomeconsistsofDNA