简介：BasedontheEigenandCrow-Kimuramodelswithasingle-peakfitnesslandscape,weproposethefitnessvaluesofallsequencetypestobeGaussiandistributedrandomvariablestoincorporatetheeffectsofthefluctuationsofthefitnesslandscapes(noiseofenvironments)andinvestigatetheconcentrationdistributionanderrorthresholdofquasispeciesbyperforminganensembleaveragewithinthistheoreticalframework.Wefindthatasmallfluctuationofthefitnesslandscapecausesonlyaslightchangeintheconcentrationdistributionanderrorthreshold,whichimpliesthattheerrorthresholdisstableagainstsmallperturbations.However,forasizablefluctuation,quitedifferentfromthepreviousdeterministicmodels,ourstatisticalresultsshowthatthetransitionfromquasi-speciestoerrorcatastropheisnotsosharp,indicatingthattheerrorthresholdislocatedwithinacertainrangeandhasashifttowardalargervalue.Ourresultsarequalitativelyinagreementwiththeexperimentaldataandprovideanewimplicationforantiviralstrategies.

简介：Evolutionarygamedynamicsinfinitesizepopulationscanbedescribedbyafitness-dependentWrightFisherprocess.Weconsidersymmetric2x2gamesinawell-mixedpopulation.Inourmodel,twoparameterstodescribethelevelofplayer’srationalityandnoiseintensityinenvironmentareintroduced.Incontrastwiththefixationprobabilitymethodthatusedinanoiselesscase,theintroducingofthenoiseintensityparametermakestheprocessanergodicMarkovprocessandbasedonthelimitdistributionoftheprocess,wecananalysistheevolutionarystablestrategy(ESS)ofthegames.WeillustratetheeffectsofthetwoparametersontheESSofgamesusingthePrisoner’sdilemmagames(PDG)andthesnowdriftgames(SG).WealsocomparetheESSofourmodelwiththatofthereplicatordynamicsininfinitesizepopulations.Theresultsaredeterminedbysimulationexperiments.