RDPManual.docx

RDPManual.docx

《RDPManual.docx》由会员分享，可在线阅读，更多相关《RDPManual.docx（109页珍藏版）》请在冰豆网上搜索。

RDPManual

UnfinishedRDP3

InstructionManual

Contents

1Introduction1

2OpeningAlignmentFiles1

3SettingAnalysisOptions1

3.1GeneralSettings1

3.2RDPSettings2

3.3GENECONVSettings2

3.4BootscanSettings3

3.5MaxChiSettings4

3.6ChimaeraSettings4

3.7SiScanSettings5

3.8LARDSettings5

3.9PhylProSettings5

3.10DNADistancePlotSettings5

3.11TOPALSettings6

3.12Recombinationratesettings7

3.13MatrixSettings7

3.14TreeSettings6

4BeginningANAnalysis6

4.1AutomatedExploratoryRecombinationAnalysis6

4.2ManualAnalyses7

5ExaminingautomatedanalysisResults7

5.1TheSchematicSequenceDisplay7

5.2TheRecombinationInformationDisplay8

5.3ThePlotDisplay8

5.4TheSequenceDisplay8

5.5TheDendrogramDisplay9

5.6TheCompatibilityMatrixDisplay9

6SavingResults9

7SupplementaryAnalyses10

8RECOMBINATIONSIGNALDETECTIONmethods10

8.1TheRDPMethod10

8.2GENECONV11

8.3Bootscan11

8.4MaxChi12

8.5Chimaera13

8.6SiScan13

8.73SEQ14

8.8PhylPro14

8.9LARD14

8.10DNADistancePlots14

8.11TOPAL15

9SuPPLEMENTARYMETHODS15

9.1BreakpointDistributionPlot（HotspotDetection）15

9.2Matrices15

10ASTEP-BY-STEPGUIDETOUSINGRDP315

10.1CompilingaGoodDataset15

10.2MakingaGoodAlignment15

10.3SettingUpaPreliminaryScanforRecombination15

10.4TestingandRefiningPreliminaryHypotheses15

9.5SoftwareCrashes/FileIncompatibilities15

11RunningRDP3fromacommandline15

12PossibleproblemsWITHUSINGRDP15

12.1PoorAlignments15

12.2RecombinantReferenceSequences15

12.3AdjacentRegionsofDifferentOrigin15

12.4Degeneracies15

12.5SoftwareCrashes/FileIncompatibilities15

13Acknowledgements15

14Appendix15

14.1ProgramFiles15

14.2CitingRDPandOtherPrograms16

15REFERENCES16

1Introduction

RDP3（RecombinationDetectionProgramversion3）isaWindows95/98/NT/XPprogramfordetectingandanalysingrecombinationsignalsinasetofalignedDNAsequences.Whileanumberofotherprogramshavebeenwrittentocarryoutthesametask（seeDrouinetal.1999,PosadaandCrandall,2001,Posada,2002,andtheweb-sitehttp:

//www.bioinf.manchester.ac.uk/recombination/programs.shtmlandreferencestherein）,mymotivationforwritingRDPhasbeentoproduceananalysistoolthatisbothaccessibletouserswhoareuncomfortablewiththeuseofUNIX/DOScommandlinesandpermitsamoreinteractiveroleintheanalysisofrecombination.

BesidesimplementingtheoriginalRDPmethoditalsoimplementsaBootscanningmethod（Salminienetal.,1995;Martinetal.,2005b）,theGENECONVmethod（Padidametal.,1999）,theMaximumChiSquaremethod（MaxChi;MaynardSmith,1992;PosadaandCrandall,2001）,theChimaeramethod（PosadaandCrandall,2001）,theSisterScanningMethod（SiScan;Gibbsetal.,2000）,the3SEQmethod（Bonietal.,2007）,theReticulatecompatibilitymatrixmethod（JakobsenandEasteal,1996）andtheTOPALDSSmethod（McGuireandWright,1998,2000）.RDP3isalsoawindowsinterfaceforarangeofotherrecombinationdetectionandanalysisprogramsincludingLARD（LikelihoodAssistedRecombinationDetection;Holmesetal.,1995）andLDHAT（McVeanetal.,20022004）.

Ifyouareimpatientandwanttostartanalysingyoursequenceswithoutreadingthemanualitisstronglyrecommendedthatyougostraighttothestep-by-stepguideinsection10.Thisguidewillhelpyouusetheprograminthewayitwasintendedtowork.

2OpeningAlignmentFiles

AnumberofdifferentalignmentfileformatsarerecognizedbyRDP3includingPHYLIP,GDE,FASTA,CLUSTAL,GCG,NEXUS,MEGAandDNAMAN.Toopenafilepressthe“Open”button（Fig1inthecommandbuttonpanel）andselectthefiletobeopened.Thedirectoryfromwhichthefileisselectedis“remembered”byRDP3whenitisshutdown.Onceloadedthealignedsequencesandtheirnamesaredisplayedinthe“sequencedisplaypanel”（Fig1）.Alsodisplayedarethedegreesofnucleotideidentityindifferentregionsofthealignedsequencesinan“identitydisplaypanel”（Fig1）.

Besidesalignmentfiles,RDP3projectfiles（witha“.rdp”extension）mayalsobeloaded.Inadditiontoalignedsequencesthesefilesalsocontaininformationonpossiblerecombinationeventsdetectedinpreviousanalysissessions.

AstillexperimentalfeaturethathasbeenimplementedinRDP3,istheabilitytoreadORFpositionsandnamesfromGenBankfiles.Theprogramrequires

（1）thataGenBankfilebeopenedafteranalignmentfileand

（2）thatoneofthesequenceswithinthemultiplealignmentisthesameasthesequenceintheGenBankfile.RDP3willautomaticallyscanthesequencesinthealignmenttocheckwhetheranymatchthesequenceintheGenBankfile.IfyouareunabletoloadaparticularGenBankfilesuccessfully,sendmethefiletogetherwithyouralignmentandI’llfixtheproblemforyou.

3SettingAnalysisOptions

Pressingthe“Options”buttoninthecommandbuttonpanelwillallowyoutomodifyRDP3ssettings.ForcasualusersofRDP3,theprogram’sdefaultsettingsshouldworkfineformostdatasets.Theonlysettingsthatyoushouldeverneedtochangeareitalicised.

3.1GeneralSettings

3.1.1Generalrecombinationdetectionoptions.Thevariousrecombinationdetectionmethodscanbesettoperceivesequencesasbeingeitherlinearorcircular.Notethatevenlinearsequencescanbeanalysedasthoughtheyarecircularandthiswillinnowayinvalidatetheanalysisresults.

ThehighestacceptableP-valuesettingisthehighestacceptableprobabilitythatsequencescouldsharehighidentitiesinpotentiallyrecombinantregionsduetochancealone（thecalculationofP-valuesdiffersforthedifferentmethodsandwillbediscussedinsection8）.TheoptimalhighestP-valuesettingvariesdependingonthenumberofsequencesinthealignmentbeinganalysed,therecombinationmethodsbeingusedtoexaminethealignment,thesizeoftheslidingwindowsthatareused（forRDP,Bootstcan,MaxChi,ChimaeraandSiScan）,andonwhetherthemultiplecomparisoncorrectionsettingisonoroff.

Thedefaultsettingformultiplecomparisoncorrectionis“on”asthismakesthecalculatedP-values“experiment-wide”（orglobal）ratherthan“currentlyselectedsequencetriplet/pairwide”（orlocal）estimatesofprobability.Notethattherearetwomultiplecomparisoncorrection“on”settings.Thedefaultis“Bonferronicorrection”butamodificationofthiscalled“step-downcorrection”isalsooffered.ThesecorrectionsactasP-valuemodifiersthatdecreasestheP-valuecutoffaccordingtothesizeofthedatasetbeingexamined.ForahighestacceptableP-valuesettingof0.05withmultiplecomparisoncorrection“off”youwouldexpectthatapproximately5%ofP-valuesthatarecalculatedwouldmaketheP-valuecutoffbychancealone.Foralargedatasetyouwouldthereforeexpectmanyfalsepositiveresults.ForthesameP-valuecutoffbutwithmultiplecomparisoncorrectionsetto“on”youwouldexpecttoonlyencounteronefalsepositivein~5%ofthedatasetsthatareexamined.Inmostsituations（<100sequenceswithanalysedsequencessharing>70%identity）ahighestacceptableP-valuesettingof0.05whenmultiplecomparisoncorrectionisonand0.0001whenitisoffshouldgivefewfalsepositivesbutstillenabletheidentificationofmostdetectablerecombinationevents.IfthecorrectionsettingisofftheP-valuecut-offmustbeverycarefullyselectedbasedonthenumberoffalsepositivesyouarepreparedtotolerate.Whenalargedatasetcontainingsequenceswithlowdiversity（e.g.100sequencesallsharing>95%identity）isanalyseditmayinfactbeimpossibletodetectanyrecombinationthatispresentifoneofthemultiplecomparisoncorrectionsettingsareon.Inthesecasesitmaybebesttoanalysethedatasetusingthepermutationtestsoffered（seesection3.1.2）withthemultiplecomparisoncorrectionsettingoffandaP-valuecut-offof0.001–thiswillgiveyousomeideaoftheexpectedfalsepositiverateforeachidentifiedrecombinationsignal.

3.1.2Permutationoptions:

Ifthe“numberofpermutationssetting”issettoanumberlargerthan0,RDP3willrunitsautomatedrecombinationdetectionanalysesinpermutationmode.Thisinvolvesgeneratingagroupofsimulatedrecombinationfreedatasets（thenumberthataresimulatedisspecifiedbyyouinthespaceprovided）,whicharethenanalysedbytheprogramusingtheexactsamesettingsthatitusestoanalysearealdataset.Thereareseveralwaysinwhichtheresultsfromsuchananalysiscanbeinterpreted.Firstly,ifRDP3identifiesmorerecombinationeventsintherealdatasetthanitdoesin95%ormoreofthesimulateddatasetsthenthisisequivalenttoaP-value<=0.05oftheirbeingnoevidenceofrecombinationinthedataset–i.e.youcanbemore95%ormoresurethatthereissomeevidenceofrecombinationinthedataset.Thisresultdoesnot,howevertellyouwhichofthedetectedrecombinationeventsareactualrecombinationevents–theresultsimplytellsyouthatsomeofthemareprobablyreal.SecondlyifRDPdetectsasinglerecombinationsignalintherealdatasetthathasabetterassociatedP-valuethanthebestsignalsin95%ormoreofthesimulateddatasetsthenthisistheequivalentofsayingthatthissignalhasaP-value<=0.05–i.e.thatyoucanbe95%confidentthattherecombinationeventassociatedwiththisrecombinationsignalisarealeventandnotafalsepositive.

RDP3canusetwodifferentapproachestosimulatethesequencesusedinthepermutationtest.Thesimplestinvolvesshufflingalignmentcolumnstodestroymostoftherecombinationsignalsevidentinanalignment.Whilethishasthepleasingeffectofmaintai