AdvPolymSci(2006)195:1–100 DOI10.1007/12_049 © Springer-VerlagBerlinHeidelberg2005 Publishedonline:13December2005 Computer-AidedConformation-DependentDesign ofCopolymerSequences PavelG.Khalatur1,2 ((cid:1))·AlexeiR.Khokhlov1,2,3 1InstituteofOrganoelementCompounds,RussianAcademyofSciences,117823Moscow, Russia [email protected] 2DepartmentofPolymerScience,UniversityofUlm,89069Ulm,Germany [email protected] 3PhysicsDepartment,MoscowStateUniversity,119899Moscow,Russia 1 Introduction:TwoParadigmsinSequenceDesign . . . . . . . . . . . . . . 5 2 NewSyntheticStrategiesinSequenceDesign . . . . . . . . . . . . . . . . 8 2.1 PreliminaryRemarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 CDSDViaPolymer-AnalogousModification . . . . . . . . . . . . . . . . . 10 2.2.1 Protein-likeCopolymers:StructureDictatesSequence . . . . . . . . . . . . 10 2.2.2 Long-RangeCorrelationsandTheirMeasure . . . . . . . . . . . . . . . . . 14 2.2.3 Hydrophobic ModificationofHydrophilicPolymers . . . . . . . . . . . . . 19 2.2.4 Adsorption-TunedCopolymers. . . . . . . . . . . . . . . . . . . . . . . . . 23 2.2.5 DesignasaSimulationofEvolutionaryProcess . . . . . . . . . . . . . . . 25 2.3 CDSDViaCopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . 31 2.3.1 ConditionsforCDSD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 2.3.2 CopolymerizationwithSimultaneousGlobuleFormation . . . . . . . . . . 32 2.3.3 EmulsionCopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.3.4 CopolymerizationNearaSelectivelyAdsorbingSurface . . . . . . . . . . . 39 2.3.5 CopolymerizationNearaPatternedSurface. . . . . . . . . . . . . . . . . . 43 2.4 DesignofMonomericUnits . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.4.1 AmphiphilicPolymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 2.4.2 HAModel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 3 PropertiesofDesignedCopolymers . . . . . . . . . . . . . . . . . . . . . . 51 3.1 SingleChains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 3.1.1 Coil-to-GlobuleTransition . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 3.1.2 KineticsoftheCollapseTransition. . . . . . . . . . . . . . . . . . . . . . . 54 3.2 PhaseBehavior:TheSequence-Assembly Problem . . . . . . . . . . . . . . 57 3.2.1 ThePolymerRISMTheory . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.2.2 Field-TheoreticCalculation . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 3.2.3 MolecularDynamicsSimulation . . . . . . . . . . . . . . . . . . . . . . . . 64 3.2.4 EvolutionaryApproach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 3.3 ChargedHydrophobic Copolymers . . . . . . . . . . . . . . . . . . . . . . 70 3.3.1 SolutionProperties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 3.3.2 DesignedCopolymersinthePresenceofMonovalentCounterions . . . . . 72 3.3.3 EffectofMultivalentCounterions . . . . . . . . . . . . . . . . . . . . . . . 74 3.3.4 StabilizationMechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 3.3.5 ExperimentalResults . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 2 P.G.Khalatur·A.R.Khokhlov 3.4 Hydrophobic-Amphiphilic Copolymers . . . . . . . . . . . . . . . . . . . . 79 3.4.1 SingleAmphiphilicChains . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 3.4.2 Coil-GlobuleTransitionVersusAggregation. . . . . . . . . . . . . . . . . . 86 3.5 AdsorptionSelectivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 3.5.1 Adsorption-TunedCopolymers. . . . . . . . . . . . . . . . . . . . . . . . . 90 3.5.2 MolecularDispenser . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Abstract Asurveyisgivenofthesimulationmethodsasappliedtothedesignofnontriv- ial sequences in synthetic copolymers aimedatachieving desired functional properties. Weconsiderarecentlydevelopedapproach,calledconformation-dependentsequencede- sign(CDSD),whichisbasedontheassumptionthatacopolymerobtainedundercertain preparation conditions is able to “remember” features of the original conformation in whichitwasbuiltandtostorethecorresponding informationintheresultingsequence. The emphasis is on copolymer sequences exhibiting large-scale compositional hetero- geneities and long-range statistical correlations between monomer units. Several new syntheticstrategiesandpolymerizationprocessesthatallowsynthesisofcopolymerswith a broad variationof their sequence distributions arereported. Wedemonstrate thatthe CDSD polymer-analogoustransformationisa versatileapproachallowingvariousfunc- tional copolymers to be obtained. Another synthetic strategy is the CDSD step growth copolymerization which is carried out under special conditions. It includes the intrin- sic possibilities of exploiting the heterogeneities of the reaction system to control the chemicalmicrostructureofthesynthesizedcopolymers,makingpossiblenewparadigms for synthesis andproductionof polymericmaterials.Inboth cases,wetrytoshow how the preparationconditions dictatecopolymer sequences. Also,wediscuss advancesthat haverecentlybeenachievedinthecomputersimulationandtheoreticalunderstandingof designed copolymers in solution and in bulk. The focus is on amphiphilic protein-like copolymers andonhydrophobic polyelectrolytes. Here, wedemonstratehow copolymer sequencedictatesstructureandproperties. Keywords Chargedheteropolymers·Copolymers·Phasebehavior·Polyamphiphiles· Sequencedesign·Simulation·Solutionproperties Abbreviations A amphiphilicgroup(monomer,monomericunit) α2 chainexpansionfactor [AM] molefractionofintermolecularaggregatesofsizeM ATC adsorption-tunedcopolymer ATRP atomtransferradicalpolymerization b bondlength C symmetricmatrixofdirectcorrelationfunctions χ Flory–Hugginsinteractionparameter (cid:1)χ effectiveinteractionparameter Cα0 bulkconcentrationofmonomerspeciesα Cα(r) instantlocalconcentrationofmonomerspeciesα Cα(z) equilibriumconcentrationprofileofmonomerspeciesα c(r) directsite-sitepaircorrelationfunction Conformation-DependentSequenceDesign 3 CRP controlledradicalpolymerization DCTT degenerativechaintransfertechnique CDSD conformation-dependentsequencedesign DF densityfunctional DFA detrendedfluctuationanalysis ∆G changeinassociationfreeenergy DL blocklengthdispersion Dλ dispersionwithinaslidingwindowoflengthλ ∆µs solvationfreeenergy ∆(q) determinantofmatrixintegralequation ∆T∗ changeintransitiontemperature ε energyparameter E unitdiagonalmatrix ε∗ criticaladsorptionenergy εPP attractionenergybetweenhydrophilic(polar)segments f fractionofchargedmonomers Φ volumefractionofmacromolecules ϕα averagefractionofmonomerspeciesαincopolymerchain φα volumefractionofmonomerspeciesα ϕα(i) intrachaincompositionprofileofmonomerspeciesα φα(r) volumefractionfieldofmonomerspeciesα FD(λ) detrendedlocalfluctuationswithinawindowoflengthλ f(λ) blocklengthdistributionfunction Fs sequencefreeenergy γ solvationparameter h Shannon’sentropy H hydrophobicmonomer(monomericunit) H symmetricmatrixoftotalsite-sitecorrelationfunctions HAmodel hydrophobic-amphiphilic(side-chain)model HPE hydrophobicpolyelectrolyte HPmodel hydrophobic-hydrophilic(polar) model h(r) totalsite-sitepaircorrelationfunction JS Jensen–Shannon divergencemeasure K associationequilibriumconstant kB Boltzmannconstant (cid:9) blocklength L averageblocklength λ lengthofslidingwindowalongcopolymersequence LH averagelengthofhydrophobicblocks LP averagelengthofhydrophilic(polar)blocks LRC long-rangecorrelation m averagenumberofcopolymerchainsperaggregate MAST macrophaseseparationtransition MFPT meanfirstpassagetime MIST microphaseseparationtransition N totalchainlength,numberofrepeatunits Nα numberofrepeatunitsoftypeα(α=A,B)inthechain NIPA poly(N-isopropylacrylamide) nJ numberofcrosslinks Nτ currentchainlength 4 P.G.Khalatur·A.R.Khokhlov Ωα volumeoccupiedbymonomerspeciesα ωα(r) averagechemicalpotentialfieldofmonomerspeciesα ODT order-disordertransition P hydrophilic(polar)monomer(monomericunit) PA hydrophilic-amphiphiliccopolymer p(αi) probabilitythatmonomerαislocatedattheithpositioninthechain PCF site-sitepaircorrelationfunction PEO poly(ethyleneoxide) PM probabilisticmodel PMF potentialofmeanforce PMMA poly(methylmethacrylate) PRISM polymer-reference-interaction-sitemodel PrP prionprotein PS polystyrene P(σ,T) probabilityofcopolymer/particlecomplex q wavenumber q wavevector ∗ q wavenumberofmaximuminstability(peakinthestructurefactor) q(r,s) propagator q†(r,s) conjugatepropagator R sizeofmicelle ρ monomernumberdensity ∗ r spatialscaleofmicrodomainstructure(domainsize) Rg radiusofgyration R2 mean-squaregyrationradius g app R apparentradiusofgyration g Rg(t) time-dependentradiusofgyration R2 partialmean-squaregyrationradiusofhydrophobic monomers gH R2 partialmean-squaregyrationradiusofhydrophilicmonomers gP R2gΘ mean-squaregyrationradiusofunperturbedchain RISM integralequationreference-interaction-sitemodel RPA randomphaseapproximation rs distancebetweennearestadsorptionsites s contourlengthofchain σ monomersize σα effectivesizeofmonomerspeciesα SASA solvent-accessiblesurfacearea Sα(q) partialscatteringfunctionformonomerspeciesα SCF self-consistent-field SCMF self-consistentmean-field SFRP stablefree-radicalpolymerization σp sizeof“parental”particle T absolutetemperature τ reducedtemperature Θ Florythetatemperature ∗ T temperatureofspinodalinstability Θα(iβ) chemicalcorrelator Conformation-DependentSequenceDesign 5 Tc criticaltemperatureofcounterioncondensation τD characteristicdiffusiontime τR reactiontimecharacterizingpolymerizationrate τ chainrelaxationtime rel Ts sequencedesigntemperature ∗ T criticalsequencedesigntemperature s u(r) site-sitepotentialdescribinginteractionbetweennonbondedmonomerunits v probabilityoflocationofterminalreactivesiteinagivenvolume W matrixofintramolecularcorrelationfunctions w(q) single-chainform-factor w(r) intramolecularsite-sitecorrelationfunction W(r) radialdistributionfunctionofmonomericunits ∗ W(r ) distributionofdomainsizes z counterioncharge(valence) 1 Introduction:TwoParadigmsinSequenceDesign Copolymers have been studied extensively for several decades, partly be- cause of their industrial and biological importance, and partly because of their interesting and sometimes perplexing properties. Many physical and mechanicalpropertiesofcopolymers,whichcomprisetwoormorecovalently bonded sequences of chemically distinct monomer species, depend on both the comonomer composition and the arrangement of these comonomers in the polymer chain. There may be significant differences, for example, be- tween twopolymer systems withthesamechemicalcomposition,butoneof whichhasthecomonomersrandomlydistributedinthechainwhiletheother haslongblocksofeachmonomertype. One may say that in many cases, sequence dictates structure and proper- ties.Toillustratethis,wewillmentiononlytwofamiliarexamples. Synthetic block copolymers can spontaneously self-assemble into highly ordered patterns of supramolecular structures (condensed modulated pha- ses), showing a surprisingly rich morphological behavior. These modulated phaseswithlengthscalesontheorderof1to103nmcanpotentiallyformthe basis for various nanotechnology applications, including the design of syn- thetic hierarchical materials, and may be effectively controlled by changing thelengthsofblocksortheirdistributionalongthechain[1]. Typically, proteins foldto organize avery specific globular conformation, knownastheprotein’snativestate,whichisingeneralreasonablystableand unique. Itisthiswell-defined three-dimensional conformationofapolypep- tidechainthatdeterminesthemacroscopicpropertiesandfunctionofapro- tein. Thefoldingmechanism andbiologicalfunctionalityaredirectlyrelated to the polypeptide sequence; a completely random amino acid sequence is unlikely to form a functional structure. In this view, polypeptide sequence 6 P.G.Khalatur·A.R.Khokhlov forces a protein to be more than a collapsed heteropolymer, but rather to assume a highly specific three-dimensional structure. Hence, a fundamen- tal issue is how functional protein sequences, which determine biologically activestructures,differfromrandomsequences.Understandingtherelation- ship between a protein’s sequence and its native structure is one of the key problemsinmodernscience[2–4]. In recent years we have seen intense interest in developing new types of functional polymer macromolecules via clever design of sequences of monomeric units in a copolymer chain. Broadly speaking, sequence de- sign may be defined as an approach aimed at finding the optimum se- quence that provides desired properties of the resultant polymer. This re- quires a scoring function which may typically be based on physical princi- ples, knowledge-based approaches, or a specifically designed function. Alas, insofar as the terms “sequence design” and “sequence engineering” imply a rational, planned approach to the creation or modification of copolymer structureandfunction,bothstillremainbeyondourcapabilitiesinageneral way. Therearetwomainparadigmsinthesequencedesignproblem. In protein science, the de novosequence designproblem consistsof find- ing asequence ofaminoacidsthatfoldintoatarget globularstructure. This problemissometimes calledtheinverse protein foldingproblem. Manycur- rent methods for de novo protein sequence design consist of numerically mutating a sequence until a maximum stability is achieved for the target structure that is usually considered as a ground state. There are a num- ber of reviews that cover this subject [5–9]. In polymer chemistry and physics, emphasis is on the development of new methods of synthesis, on the controlof(co)polymer stereochemistry and architecture,and onthede- sign of high performance polymeric materials tailored for specific uses and properties. Thedifferencebetweenthetwosequencedesignconceptsisrelatedtosev- eral essential differences between natural and non-natural copolymers. We mentionhereonlyafewofthem. The order of amino acids in a polypeptide chain produced by the syn- thetic apparatus of the living cell is always the same for a particular protein so that all the protein sequences of a given type are structurally identical copies inevery cellinaliving organism. We cannotdistinguish oneindivid- ualproteinsequencefromanother.Formostsyntheticcopolymers,produced industrially or synthesized in research labs, the occurrence of a certain de- gree ofsequence disorder is almost inevitable. Therefore, if we are speaking aboutasynthetic copolymersequence, wemean, explicitlyorimplicitly,that averaging over many different sequences has been carriedout.Foraprotein to function, it must be in its highly specific native conformation that is sta- ble only in a narrow temperature region. On the other hand, the properties andfunctionsofsyntheticcopolymersarenotsotightlyrelatedtotheircon- Conformation-DependentSequenceDesign 7 formation. Moreover, we are mainly interested in the nonunique copolymer conformations. The interior of proteins has the packing density of a mo- lecular crystal while synthetic globules are typically liquid-like. This list is of course far from exhaustive, but should rather be taken as simply a set of examples. In the present article we will deal mainly with synthetic macromolecules andpracticallywillnottouchonbiopolymers. Diblock and repeated-block AB copolymers are the simplest examples of two-letter copolymers made up of two different monomer species, denoted by letters A and B. More sophisticated distributions of chemically differ- ent groups along the chain are characteristic of random and random-block copolymers, including uncorrelated or ideal random copolymers and so- calledcorrelatedrandomcopolymers.Intheformerclass,thecorresponding chemical sequences are uncorrelated and this corresponds to Bernoulli or zeroth-orderMarkovprocesses[10]. Inthelatterclass,thecorrelationinthe sequences of both types of A and B segments is defined by means of a first- order Markov process. It is important to emphasize that in both cases the correlationscharacterizingthedistributionofmonomericunitsalongcopoly- mer chains decay exponentially. There are, however, copolymers for which thisisnotthecase.Inthisreview,wewillconsiderjustthesecopolymers,fo- cusing on the computer-aided design of their chemical sequences as well as onthepropertiesofdesignedpolymers. Althoughrecentyearshavewitnessedanimpressiveconfluence ofexperi- ments and statistical theories, presently there is no comprehensive under- standingoftheinterrelationbetweenchemicalsequencesinsyntheticcopoly- mers and the conditions of synthesis. One has merely to glance at recent literature in polymer science and biophysics to realize that the problem of sequence-property relationship isby no means entirely solved. As always, in thesecircumstances,analternativetoanalyticaltheoriesiscomputersimula- tions,whicharedesignedtoobtainanumericalanswerwithoutknowledgeof ananalyticalsolution. Thecomputersimulationsarelikelytobeusefulintwodistinctsituations— thefirstinwhichnumericaldataofaspecifiedaccuracyarerequired,possibly forsomeutilitarianpurpose;thesecond,perhapsmorefundamental,inpro- viding guidance to thetheoretician’s intuition, e.g., by comparing numerical resultswiththosefromapproximateanalyticalapproaches.Asaconsequence, the physical content of the model will depend upon the purpose of the cal- culation. Our attention here will be focused largely on the coarse-grained (lattice and off-lattice) models of polymers. Naturally, these models should reflect those generic properties of polymers that are the result of the chain- likestructureofmacromolecules. Apartfromtheintroductorysection,thearticleissubdividedintotwoma- jorsections:SynthesisandPropertiesofdesignedcopolymers. 8 P.G.Khalatur·A.R.Khokhlov 2 NewSyntheticStrategiesinSequenceDesign 2.1 PreliminaryRemarks Today, the majority of all polymeric materials is produced using the free- radical polymerization technique [11–17]. Unfortunately, however, in con- ventional free-radical copolymerization, control of the incorporation of monomer species into a copolymer chain is practically impossible. Fur- thermore, in this process, the propagating macroradicals usually attach monomeric units in a random way, governed by the relative reactivities of polymerizing comonomers. This lack of control confines the versatility of the free-radical process, because the microscopic polymer properties, such as chemical composition distribution and tacticity are key parameters that determinethemacroscopicbehavioroftheresultantproduct. The absence of control of the incorporation of monomers into the poly- meric chainimplies that many macroscopic propertiescannot be influenced toalargeextent.Therefore,inrecentyearsmuchefforthasbeendirectedto- wards the development of controlledradical polymerization (CRP) methods forthepreparationofvariouscopolymers(forarecentreview,see[17]). These methods are based on the idea of establishing equilibrium be- tween the active and dormant species in solution phase. In particular, the methods include three major techniques called stable free-radical polymer- ization(SFRP),atomtransferradicalpolymerization(ATRP),andthedegen- erative chaintransfer technique (DCTT) [17]. Althoughsuchsyntheses pose significant technical problems, these difficulties have all been successively overcomeinthelastfewyears.Nevertheless, theprocedureofpreparationof theresultingcopolymersremainssomewhatcomplicated. On the other hand, it should be realized that radical copolymerization at heterogeneous conditions offers additional unique opportunities not avail- ableinhomogeneous(solution)copolymerization.Theseincludetheintrinsic possibilitiesofexploitingtheheterogeneitiesofthereactionsystemtocontrol thechemicalmicrostructureofthesynthesizedcopolymers,makingpossible new paradigms for synthesis and production of polymeric materials. In this contribution, wediscusssomenewsynthetic strategies, whichhavebeen de- velopedinrecentyearstoprovideeffectivecontrolofthechemicalsequences. Inaseriesofpublications[18–20],theconceptofconformation-dependent sequence design (CDSD) of functional copolymers has been introduced (for recent reviews, see [21–25]). The essence ofthe proposedapproachisbased ontheassumptionthatacopolymerobtainedundercertainpreparationcon- ditions isableto remember features oftheoriginal(“parent”) conformation in which it was built and to store the corresponding information in the re- sulting chemical sequence. In other words, this concept takes into account Conformation-DependentSequenceDesign 9 astrongcouplingbetweentheconformationandprimarystructureofcopoly- mers during their synthesis. Ideologically, the approach[18–20] bears some similarities withthatproposed earlier in thecontextoftheproblemsofpro- tein physics [26–30], however, it is aimed at synthetic copolymers rather than biopolymers. The original idea for protein design [26–30] consisted of running through sequences of amino acids to determine which sequence (or sequences) had the lowest energy in a unique (target) conformation. In the Introduction, we stressed the differences between this approach and CDSD. In polymer chemistry, there are two known CDSD techniques: (i) the chemical modification (polymer-analogous transformation) of homopoly- mers and (ii) the step-growthcopolymerization ofmonomers withdifferent propertiesunderspecialconditions.Wewilladdressboththesetechniques. Polymer simulations are being done on different levels. An atomistic modelofapolymercontainsalltheatomsthatarepresentintherealpolymer. Coarse-grained models simplify the problem by combining atoms into ef- fectiveunitedatoms.Inthisway,onlysignificantmicroscopicinformation— “essential features” of the real system —is retained. The unit can represent a chemical groupof a few atoms, a monomeric unit in a polymer, groupsof monomeric units, or chain segments of various lengths. Certainly, whether acoarse-graineddescriptionisadequateforunderstandingaparticularpoly- mersystemdependsverymuchontheproblembeingstudied.Thechallenge lies in selecting just the right amount of atomistic detail to build coarse- grainedmodelswithmaximumgenerality.Toacertaindegree,ofcourse,that ispreciselywhatphysicsisabout.Therefore,thequestionishowcanwecon- structcopolymermodelsthataresufficientlyrealistictocapturetheessential features ofrealmacromoleculesyet simple enough toallowlarge-scalecom- putationsofpolymerconformationanddynamics? Therearemanykindsofpolymerizingmonomersusedtomakeupcopoly- mers. These differ in physical and chemical properties. One of the most importantdifferences(essentialfeatures)istheirsolubility,thatis,howmuch they like or dislike a solvent, e.g., water. Hence the chemical and atomistic details of different monomeric units may not be necessary to understand the properties of many “two-letter” copolymers. In what follows, we will mainly use the so-called HP model [31]. This two-letter model of a lin- ear hydrophobic/hydrophilicmacromoleculereflectsthespirit ofminimalist models,inthatitissimpleyetbasedonaphysicalprinciple. The HP model is a coarse-grained (lattice or off-lattice) polymer model thatabstractsfromrealpolymersintwoimportantways:(i)Insteadofmod- eling the positions of all atoms of the polymer, it models only the backbone structure of the polymer, i.e., one position for each monomeric unit. (ii) Usually, only the hydrophobic interaction between the monomeric units is modeled, therefore the model distinguishes only two kinds of monomeric units,namelyhydrophobic(H)andhydrophilic(orpolar,P). 10 P.G.Khalatur·A.R.Khokhlov 2.2 CDSDViaPolymer-AnalogousModification 2.2.1 Protein-likeCopolymers:StructureDictatesSequence The studies of structures formed by copolymers consisting of two kinds of monomeric units constitute a rather large field of polymer chemistry and physics [32]. The systems that are most extensively studied are block- copolymers (withablockprimary structure) andrandom copolymers (with statisticalprimarystructure).Sometimes,copolymerswithsomeshort-range correlations along the chain are also investigated. Such correlations will al- ways show up after the copolymerization process, if the probability of add- ition of unit A or B to the growing chain depends on the type of unit that was added on the previous polymerization step [17]. The type of primary structure that emerges in this case can be characterized as “random with short-range correlations”. On the other hand, globular proteins can also be regarded from a very rough viewpoint as a kind of binary copolymer. In- deed,themostimportantdifferencebetweenthemonomericunitsofglobular proteins is that some amino acid residues are hydrophobic, while others are hydrophilic or charged. We can very roughly attribute the index H to the former type of units and the index P to the latter ones [33]. If we then ana- lyzetheprimarystructuresoftheglobularproteinsobtainedinthiswayand compare them with the simple primary structure of conventional synthetic copolymers, we shoulddrawthe conclusionthat protein-originated ABtexts aremuchmoreinformativeandspecific. It is generally believed that in globular proteins the hydrophilic P units mainly cover the surface of the globule, giving rise to their stability against intermolecular aggregation, while hydrophobic H units main form the core of theglobule [33]. Itcan beassumed that such a requirement (in thedense globular state the P amino acids should be on the surface and the H amino acids inthecore)israther restrictive, i.e., itissatisfied onlyforavery small fraction of all possible primary structures. Moreover, since the HP correla- tions defined in such a way depend on the conformation of the globule as awhole(i.e.,ontheternarystructure),theyshouldbecharacterizedaslong- rangeones. The question is, whether such primary structures can be obtained for bi- narycopolymers, notobligatorilyofbiologicalorigin.Itiseasy todothisby computersimulation[18],andmuchmoredifficultinrealexperiments.How- ever,inbothcasesthecorrespondingprocedureshouldinvolvethefollowing stagesthatareschematicallydepictedinFig.1: Stage1. We take a homopolymer coil with excluded volume interactions in agoodsolvent.