(nsimenps.zip212kB).Addressquestionsandi
ilableforUNIX(nsimenv.ps.Z286kB)andMSDOS
nquiriestoMichaelHinesorTedCarnevaleDigi
PostScriptversionsofthismanuscriptareava
),pp.1179- CONTENTS
opyright©1 1.INTRODUCTION
ust15,1997    1.1Theproblemdomain
le,NeuralC    1.2Experimentaladvancesandquantitative
997byN.T.C    modeling
ment'byM.L 2.OVERVIEWOFNEURON
served.HTM 3.MATHEMATICALBASIS
.HinesandN    3.1Thecableequation
,Volume9,N    3.2Spatialdiscretizationinabiological
tof'TheNEU    context:sectionsandsegments
.T.Carneva    3.3Integrationmethods
chnology,a       3.3.1Efficiency
ight©1997b 4.THENEURONSIMULATIONENVIRONMENT
ytheMassac    4.1Thehocinterpreter
avigationc    4.2Aspecificexample
tituteofTe       4.2.1Firststep:establishmodel
llrightsre       topology
umber6(Aug       4.2.2Secondstep:assignanatomical
gandgraphi       andbiophysicalproperties
arnevaleam       4.2.3Thirdstep:attachstimulating
talpreprin       electrodes
ionEnviron       4.2.4Fourthstep:controlsimulation
omputation       timecourse
Lformattin    4.3Sectionvariables
csforpagen    4.4Rangevariables
1209.Copyr    4.5Specifyinggeometry:stylizedvs.3-D
husettsIns    4.6Densitymechanismsandpoint
dM.L.Hines    processes
RONSimulat    4.7Graphicalinterface
              4.8Object-orientedsyntax
                 4.8.1Neurons
                 4.8.2Networks
1.INTRODUC 5.SUMMARY
TIONNEURON ACKNOWLEDGMENTS
(Hines1984 REFERENCES
ervations.Tothetheoreticianitisameansfor
;1989;1993;1994)providesapowerfulandflex
osenthal1992)andareconstrainedtooperatew
nenetal.1995)cellularmechanismsoflearnin
thcoming.Copyright©1997bytheMassachusett
function(LyttonandSejnowski1992)presynap
eprintof'TheNEURONSimulationEnvironment'
ofspatialandtemporalscales(CarnevaleandR
nothaveanalyticalsolutions,andintuitioni
snotareliableguidetounderstandingthework
itmeansthatthechoiceofwhichdetailstoincl
xperimentaladvancesandquantitativemodeli
ragonistsandantagonistsgeneticengineerin
ionoflibrariesofidentifiedneuronsandneur
alisticquantitativemodelsthatmustbeusedt
(DestexheandSejnowski1995;Traynelisetal.
1993)mechanismsunderlyingcommunicationbe
Häusseretal.1995;HinesandShrager1991;Mai
g(Brownetal.1992;Tsaietal.1994a)cellular
iestoMichaelHinesorTedCarnevaleDigitalpr
.HTMLformattingandgraphicsforpagenavigat
reationofbiologicallyrealisticquantitati
nsthatdescribebrainmechanismsgenerallydo
tforparticularphenomena.Tothestudentinal
aboratorycourseitprovidesavehicleforillu
ysicsofreceptorsandchannels,theconstruct
lvedinperception,learning,andsensorimoto
ross-validatingdata,estimatingexperiment
overtheorderhiddenwithintheintricacyofbi
hisarticledescribestheconceptsandstrateg
ationprocessinginthebrainresultsfromthes
vesnonlinearmechanismsthatspanawiderange
ingofthecellsandcircuitsofthebrain.Furth
ermore,thesenonlinearitiesandspatiotempo
heseproblemsbyenablingboththeconvenientc
'doesnotmean'infinitelydetailed.'Instead
etionoftheinvestigatorwhoconstructsthemo
allyinaccessibleparameters,anddecidingwh
ngExperimentaladvancesdriveandsupportqua
ipleimpalementsofvisuallyidentifiedcells
onalclassesthathavebeencharacterizedanat
lly,andtheanalysisofneuronalcircuitsinvo
atcatalyzestheformulationofnewhypotheses
otestthesehypotheses.Someexamplesfromthe
996;Jaffeetal.1994)drugeffectsonneuronal
Bernanderetal.1991;CaullerandConnors1992
byM.L.HinesandN.T.Carnevale,NeuralComput
sInstituteofTechnology,allrightsreserved
vemodelsofbrainmechanismsandtheefficient
stsubsetofanatomicalandbiophysicalproper
ke,apowerfulsimulationtoolsuchasNEURONca
esioningofcellsphotoreleaseofcagedcompou
ioncopyright©1997byN.T.CarnevaleamdM.L.H
ologicalphenomena,theorderthattranscends
dthroughtheuseofsuchmodelsincludethecell
al.1996)thalamicnetworks(Destexheetal.19
lyrealisticmodelsofelectricalandchemical
signalinginneuronsandnetworksofneurons.T
verelylimited.NEURONisdesignedtoaddresst
otheexperimentalistNEURONoffersatoolforc
echanismsinasimplifiedformthatismorerobu
alsmultisiteelectricalandopticalrecordin
rintegration.Theresultisadataavalancheth
gastheempiricalbasisforthebiologicallyre
perimentalist,theoretician,andstudentali
micalandelectricalsignals(Destexheetal.1
ntitativemodeling.Overthepasttwodecadest
usmeasurementofelectricalandchemicalsign
stefficientuse.1.1TheproblemdomainInform
preadandinteractionofelectricalandchemic
ithintheintricateanatomyofneuronsandthei
rinterconnections.Consequentlytheequatio
ivemodelingtoolsthatweredevelopedwithout
simulationoftheoperationofthesemechanism
etherknownfactsaccountforexperimentalobs
stratingandexploringtheoperationofbrainm
hefieldofneurosciencehasseenstrikingdeve
lopmentsinexperimentaltechniquesthatincl
iredpre-andpostsynapticneuronssimultaneo
gofionchannelsandreceptorsanalysisofmRNA
sareresponsibleforimpressiveprogressinth
ofbrainfunction,whileatthesametimeservin
ularmechanismsthatgenerateandregulateche
ibleenvironmentforimplementingbiological
s.Inthiscontexttheterm'biologicalrealism
udeinthemodelandwhichtoomitareatthediscr
simultaneousintracellularrecordingfrompa
hysicalpropertiesfromthesameneuronphotol
edefinitionofthemolecularbiologyandbioph
largelistoftopicsthathavebeeninvestigate
tweenneuronsintegrationofsynapticinputs(
oscillations(Destexheetal.1993a;Lyttonet
testinghypothesesanddeterminingthesmalle
omically,pharmacologically,andbiophysica
eencounteredinmostnonbiologicalsystems,s
tic(LindgrenandMoore1989)andpostsynaptic
ionofthissimulator,withemphasisonthosefe
takingthesefeaturesintoconsiderationisse
tiesthatisnecessaryandsufficienttoaccoun
nbeanindispensableaidtodevelopingtheinsi
nnewdrugssuchaschannelblockersandrecepto
andbiophysicalpropertiesfromthesameneuro
ncoding(Hsuetal.1993;MainenandSejnowski1
aturesthatareparticularlyrelevanttoitsmo
otheutilityofmanyquantitativeandqualitat
del,andnotforcedbythesimulationprogram.T
stthanthetypical'wetlab'experiment.Forex
uronsinvitroandinvivousingpatchclampmult
gquantitativeanalysisofanatomicalandbiop
ndsforspatiallyprecisechemicalstimulatio
n'knockout'mutationsTheseandotheradvance
)actionpotentialinitiationandconduction(
ghtandintuitionthatisneededifoneistodisc
udehigh-qualityelectricalrecordingfromne
ation,Volume9,Number6(August15,1997),for
iesthathaveguidedthedesignandimplementat
alsignalswithinandamongneurons.Thisinvol
ralcomplexitiesarequiteunlikethosethatar
thecomplexityofaccidentandevolution.1.2E
93b;Destexheetal.1994)neuralinformatione
995;Softky1994)Addressquestionsandinquir
ines.................................PbN