Download Coevolution analysis of viral sequences: from HBV - LCQB

Post-docpositionavailableattheUniversitéPierreetMarieCurie,Paris
Coevolutionanalysisofviralsequences:fromHBVtoageneral
framework
Context&BiologicalQuestion.Aparticularfocushasbeendrawninrecentyearstocoevolvingresidues,withina
proteinandamongproteins.Coevolvingresiduesinaproteinstructure,possiblyacomplex,correspondtogroups
ofresidueswhosemutationshavearisensimultaneouslyduringtheevolutionofdifferentspecies,andthisisdueto
severalpossiblereasonsinvolvingthethree-dimensionalshapeoftheprotein:functionalinteractions,
conformationalchangesandfolding.Severalstudiesaddressedtheproblemofextractingsignalsofcoevolution
betweenresidues.Allthesemethodsprovidesetsofcoevolvedresiduesthatareusuallycloseinthethreedimensionalstructure,formconnectednetworkscoveringroughlyathirdoftheentirestructure,andhavebeen
demonstratedforafewproteincomplexestoplayacrucialroleinallostericmechanisms,tomaintainshortpaths
innetworkcommunicationandtomediatesignaling.Thesemethodsareapplicabletoproteinfamiliesdisplayinga
largenumberofevolutionarilyrelatedsequencesandsufficientdivergence,thesecharacteristicsconstitutingthe
bottleneckoftodaycoevolutionanalysismethods.Inthisprojectweshallgobeyondthisbottleneckwithafurther
developmentofBIS2,afastalgorithmforthecoevolutionanalysisofrelativelysmallsetsofsequences(where
“small”means<50sequences)displayinghighsimilarity[1,2,4].BIS2canbeappliedtomanyproteins,
characteristicofvertebrateorviralspecies,whereothercoevolutionmethodscannotbeappliedbecauseofthe
reducednumberofsequences(eithercomingfromspeciesorfrompopulations)andtheirconservation.
Inthisproject,weaimatdecryptingmechanisticfeaturesofthehepatitisBvirus(HBV)fusionprocessandat
proposinganexperimentallysupportedmodelforHBVmembranefusionmechanism.Thiswillbedonethrougha
multidisciplinaryapproach,combiningbioinformaticsanalysisofmolecularcoevolutionandexperimentalassays.
DuetothestudyofsequencescomingfromHBviralpopulations(identifiedingeographicalregions,orevenin
singlepatients),
Previousstudies.TheBISmethodwasrecentlyusedtoaddresstwoquestionsonhepatitisCvirus(HCV).Thefirst
concernsthereconstructionoftheHCVprotein-proteininteractionnetwork[2]andthesecondtheunderstanding
oftheHCVfusionmechanism[3].Thesetwoquestionscouldnothavebeenaddressedbyothercurrentstate-ofthe-artcoevolutionmethods,duetotheirrequirementforlargenumbersofsequenceswithhighdivergence.This
pioneeringworkconstitutesthefirstcoevolutionanalysisofanentireviralgenomeandopensthewayto
coevolution-basedreconstructionofprotein-proteininteractionnetworksinviruses.
Directionsdevelopedbythepost-doc.Withintheproject,thespecificobjectivesofthepostdocare:1.the
developmentofacoevolutionanalysistoolforHBVviralpopulations;2.thecomputationalidentificationofthe
residuesbeinginvolvedinmembranefusioninHBV.
1. ThedevelopmentofacoevolutionanalysistoolforHBVviralpopulations.
WeproposetoextendtheBIS2methodinseveraldirectionstakingintoconsiderationthenonstandardgenomesof
viral species and viral populations. Among the several aspects that we will explicitly handle, we highlight a
systematic inclusion of independent studies of several genotypes in the predictions of intra and inter-protein
signals, a nucleotide analysis of the viral sequences leading to a distinction of protein functions and protein
interactions,viralgeneoverlappingandgenomerecombination.Wewilldevelopanimprovedmethod,specifically
conceivedforviralsequences,thatleadstothepredictionofcoevolvingresidueswithhighstatisticalsignificance.
Suitable statistical scores (p-value), taking into consideration the various constraints that we shall include in our
study,shouldbedefined.
2.ThecomputationalidentificationoftheresiduesbeinginvolvedinHBVmembranefusion.Weshallextendthe
useofourcoevolutionanalysistool(seeObjective1),topredictintra-andinter-proteininteractionsunderlyingthe
HBV entry process, including the binding and membrane fusion mechanisms. We shall tackle the problem of the
interactions among surface proteins, even for proteins where no structure is available. We shall reconstruct
coevolutionsignalsonindependentanalysisofgenotypesequencefamiliesbasedonwhichweshallidentifypairs
of physically proximal coevolved clusters of residues and use them to possibly support hypotheses of large
collectivemovements(drivingconformationalchanges).Dependingontheevolutionoftheproject,wemightbe
abletoextendtheanalysistodifferentactorsinthefusionprocess.
Theprojectdemandsthedevelopmentofefficientalgorithms.
References
1 Dib,L.,Carbone,A.Proteinfragments:Functionalandstructuralrolesoftheircoevolutionnetworks.PLoSONE7(11),
e48124(2012).
2 Champeimont,E.,Laine,E.,Hu,S.-W.,Penin,F.&Carbone,A.CoevolutionanalysisofHepatitisCvirusgenometoidentify
thestructuralandfunctionaldependencynetworkofviralproteins.ScientificReports6:26401(2016).
3 Douam,F.etal.Aproteincoevolutionmethoddesignedforconservedsequencesuncoverscriticalfeaturesoftheoriginal
HCVfusionmechanismandprovidesmolecularbasisforthedesignofeffectiveantiviralstrategies.Inrevision(2016).
4 Baussand,J.&Carbone,A.Acombinatorialapproachtodetectco-evolvedamino-acidnetworksinproteinfamilieswith
variabledivergence.PLoSComputationalBiology5,e1000488,(2009).
Environment:
ThepostdocwillbenefitfromtheinteractionwiththeexperimentalvirologistsattheEcoleNormaleSupérieurede
Lyon(Francois-LoicCossetLab–CentreInternationaldeRechercheenInfectiologie;http://ciri.inserm.fr/leciri/presentation/#)toincludeintheformalmodelallaspectsofthecomplexityofHBVgenomethatwillbe
necessarytoobtainaccuratepredictions.Afundamentalaspectoftheprojectistheback-and-forthbetweenthe
experimentalistsandthecomputerscientists/mathematiciansinvolved;itwillbefundamentaltodesignthe
experimentalstrategiesbasedonanappropriateunderstandingofgenomiccoevolutionsignalsandtoconceivean
appropriateformaltoolforthecoevolutionanalysisofviralgenomes.Therearetwodistinguishedoutputsforthe
project:theidentificationofthemechanisticfeaturesoftheHBVfusionprocessandthenoveltoolforcoevolution
analysisofviralgenomesthatwillbemadeavailabletothecommunityforexploitationwithotherviralgenomes.
Theresultsobtainedbythepostdocwillbemadeavailabletothecommunity.Theywillbeofinterestfortheentire
virologycommunitybecausewedoexpectthemtobeapplicabletootherviruses.
Theresearchteam(AnalyticalGenomicsteam)ispartoftheLaboratoryofComputationalandQuantitativeBiology
(LCQB)UMR7238CNRS-UPMC,headedbyA.Carbone.TheLCQBhasaverydynamicenvironmentwhere8teams
work,indifferentarea,atthefrontierofcomputationalbiology.Informationonthelabandtheteamscanbe
foundathttp://www.lcqb.upmc.fr.Fortheteam,seehttp://www.lcqb.upmc.fr/AnalGenom/and
http://www.lcqb.upmc.fr/AnalGenom/projects.html
Period:Thepostdocwilllast2years.ItisavailablefromMarch1st,2017.ItisfundedbyANRS(AgenceNationale
pourlaRecherchecontreleSIDAetlesHépatitesVirales).
Postdocprofile.WelookforapostdocwithaverysuccessfulPhDexperienceandanexcellentpublicationrecordin
either
1.sequenceanalysisandevolution.
or
2.mathematics/computersciencewithknowledgeingraphtheoryandcombinatorics,patternmatching,andin
proteinstructures.
Contacts:
A.Carbone–[email protected]
LaboratoryofComputationalandQuantitativeBiology(LCQB)UMR7238CNRS-UPMC
UniversitéPierreetMarieCurie,Paris.