behavioriyunssr
iyunssr 时间:2021-04-09 阅读:()
RESEARCHARTICLEOpenAccessGenome-wideSSR-basedassociationmappingforfiberqualityinnation-wideuplandcottoninbreedcultivarsinChinaXinhuiNie1,2,CongHuang1,ChunyuanYou2,WuLi3,WenxiaZhao1,ChaoShen1,BeibeiZhang1,HantaoWang1,ZhenhuaYan1,BaoshenDai1,MaojunWang1,XianlongZhang1andZhongxuLin1*AbstractBackground:SinceuplandcottonwasintroducedintoChinaduringthe1920s–1950s,hundredsofinbreedcultivarshavebeendeveloped.
Toexplorethemoleculardiversity,populationstructureandelitealleles,503inbredcultivarsdevelopedinChinaandsomeforeigncultivarsfromtheUnitedStatesandtheSovietUnionwerecollectedandanalyzedby494genome-wideSSRs(SimpleSequenceRepeats).
Methods:Fourhundredandninety-fourpairsofSSRswithhighpolymorphismanduniformdistributionon26chromosomeswereusedtoscanpolymorphismsin503nation-wideuplandcottons.
TheprogramminglanguageRwasusedtomakeboxplotsforthephenotypictraitsindifferentenvironments.
Molecularmarkerdataand6fiberqualitytraitswereanalyzedbythemethodofMLM(mixedlinearmodel)(P+G+Q+K)intheTASSELsoftwarepackageonthebasisofthepopulationstructureandlinkagedisequilibriumanalysis.
Thelociofeliteallelicvariationandtypicalmaterialscarryingelitealleleswereidentifiedbasedonphenotypiceffectvalues.
Results:Atotalof179markerswerepolymorphicandgenerated426alleleloci;thepopulationbasedonmoleculardiversitywasclassifiedintosevensubpopulationscorrespondingtopedigreeorigin,ecologicalandgeographicaldistribution.
Theattenuationdistanceoflinkagedisequilibriumdroppedsignificantlyupto0–5cM.
Associationmappingforfiberqualityshowedthat216markerlociwereassociatedwithfiberqualitytraits(PFE(0.
581)>FS(0.
474)>MV(0.
445)>FU(0.
410)>SF(0.
380).
Itwasmoreimportanttofurtheranalyzeonetraitbetweentwoenvironments;takingFig.
1PCAplotsof503uplandcottoncultivarsbasedonSSRs.
Bluetriangles,greensquares,browntriangles,purplediamonds,lightbluesquare,reddiamondsandyellowcirclesrepresentcultivarsfromdifferentregionsFig.
2PopulationstructureanalysisandlineschartoftheKvaluewithLnP(D)valueandΔKvaluebasedonstructureanalysis.
LinechartoftheLnP(D)valuewiththechangeofK(a);LinechartoftheΔKvaluewiththechangeofK(b);Populationstructureof503cultivar-basedSSRmarkers(c).
Q-plotshowingtheclusteringof503uplandcottoncultivarsbasedontheanalysisofgenotypicdatausingSTRUCTURE.
Eachcultivarisrepresentedbyaverticalbar.
Thecoloredsubsectionswithineachverticalbarindicatethemembershipcoefficient(Q)ofthecultivartodifferentclustersNieetal.
BMCGenomics(2016)17:352Page4of16FUHMLasexample(Fig.
5a),thecorrelationsrangedfrom0.
40to0.
76,andthecorrelationwas0.
76forFUHML_12KELandFUHML_13KEL.
Theredlinethatwasnearthe45°asthelineofgreatestslopein-dicatedacorrelationbetweenFUHML_12KELandFUHML_13KEL.
Associationmappingoffiberquality-relatedtraitsBasedonthegenotypedata,thePCAmatrix,thekinshipmatrix,andthefiberqualitytraitsdataoftheBLUPre-sultsin8environments,amixedlinearmodelwasusedtoanalyzethemarker-traitassociations.
Duringassoci-ationmapping,threemodels,GLM(P+G)+Q,GLMFig.
3ThedistributionofLDamong426SSRlocion26chromosomesin503uplandcottoncultivars(R2value).
SSRmarkerswerealongtheX-axis.
Eachpixelabovethediagonalrepresentsther2sizeofthecorrespondingpairsofmarkers,asshowninthecolorcodeattheupperright,andeachpixelbelowthediagonalrepresentstheP-valuesizeoftestingtheLDatthelowerrightFig.
4TheboxplotsofthechangingtrendsofsixfiberqualitytraitsineightenvironmentsNieetal.
BMCGenomics(2016)17:352Page5of16(P+G)+PCs,andMLM(G+P+Q+K),werecom-paredwitheachotherintheassociationanalysis(Additionalfile9).
ThecontroleffectofthepopulationstructureforFUHML,FS,SF,andFEinthethreemodelswassimilar.
However,theMVanalysisusingtheMLM-Q-Kmodelwassuperiortotheothertwomodels,andtheFUanalysisusingtheGLM-QandMLM-Q-KmodelswasbetterthantheofGLM-PCAmodel.
Accordingtotheabovecomparisonresults,theMLM-Q-Kmodelhadbetterperformance.
Atotalof179SSRmarkerswereusedformarker-traitassociationafterfilteringfor5%minimumalleles,amongwhich91(50.
84%)markerswereassociatedwithfiberqualitytraitsattheP<0.
05level.
FifteenmarkersweresignificantlyassociatedattheP<0.
01level(Fig.
6).
Anaverageof3.
5markerswasdetectedoneachchromosome(rangingfrom1to8),withthemaximumof8markersonChr01andChr19.
Onemarkerwasgenerallyassociatedwithseveraltraits.
Forexample,MON_DC40013onChr01wasrelatedtoFUHML,SF,FUandFS,andNAU2564onChr07wasrelatedtoFUHML,SFandFS.
Therewere216lociassociatedwithfiberqualitycom-ponentsattheP<0.
05significancelevel,amongwhich27weresignificantattheP<0.
01level.
Therangeofphenotypicvariationexplanation(PVE)observedwasfrom0.
58%(MON_DPL0042b)to5.
12(NAU3084c),withanaverageof2.
70%(Additionalfile10).
Amongthe6traits,FSwasassociatedwiththemostloci,uptoamaximumof61(P<0.
05)and7(P<0.
01);PVErangedfrom0.
58(MON-DPL0042b,P<0.
05)to3.
17%(NAU2836a,P<0.
001),withameanof2.
63%.
TheremarkablecontributionlociwereNAU2858a(2.
92%),BNL3089a(2.
86%)andMONCGR5399c(2.
73%),especiallyattheP<0.
001level.
FUHMLwasassociatedwiththesecondnumberofloci,uptoamaximumof46(P<0.
05)and4(P<0.
01);PVErangedfrom0.
59(NAU3092a,P<0.
05)to3.
11%(NAU5480b,P<0.
001),withameanof2.
35%,inwhichNAU5480a(2.
75%)hadasignificantcontribu-tionatP<0.
001.
Therewereupto42(P<0.
05)and6(P<0.
01)lociassociatedwithSF,andPVEsrangedfrom0.
68(HAU2835b,P<0.
05)to5.
12%(NAU3084c,P<0.
001),Fig.
5ThecorrelationsofsixfiberqualitytraitsineightenvironmentsNieetal.
BMCGenomics(2016)17:352Page6of16Fig.
6Thedistributionofthelocatedmarkersassociatedwiththefiberquality-relatedtraitson26chromosomesNieetal.
BMCGenomics(2016)17:352Page7of16withameanof3.
80%.
NAU3084b(5.
08%),MON-DPL0893a(4.
43%),andMON-DC40013b(3.
03%)con-tributedprominentlytoSFatP<0.
001.
Therewere25(P<0.
05)and3(P<0.
01)lociassociatedwithFE,withPVErangingfrom0.
70(MON-CGR5423b,P<0.
05)to2.
99%(HAU4806-SSCPa,P<0.
001).
BNL846bandMON_CGR5113bcontributedtoFEatP<0.
01.
Therewere23(onlyatP<0.
05)lociassociatedwithMV,withPVErangingfrom0.
66(NAU3138b)to1.
46%(DPL0457a);NBRI-HQ524733bcontributingtoMVwasdetectedatP<0.
05.
Therewere19(P<0.
05)and7(P<0.
01)lociassociatedwithFU,andthePVEsrangedfrom1.
06(HAU1279d)to2.
51%(HAU1166a),withameanof2.
14%.
ThecontributionlociwereobservedinMON-CGR5602a(2.
39%),MON-DPL0893a(2.
26%),MON-DC40013b(2.
21%),NAU3084c(1.
93%),HAU1166b(1.
87%)andNAU3084b(1.
82%)atP<0.
01.
Exploringeliteallele-relatedgenesinthecottongenomeThereferencesequencesof91eliteallelelociassociatedwithfiberqualitytraitswereexploredbasedonrelatedgenesinG.
arboreum,G.
raimondiiandG.
hirsutum,.
Threeallelicvariationlociwererelatedtogenefunc-tionalannotationoffiberqualitytraitsinG.
arboreum(Additionalfile11).
HAU0211wasassociatedwithFUandSFonChr12;itshomologousgenesinG.
arboreumandArabidopsisthalianawereCotton_A_01461andAT5G16560.
1,respectively,whichwereannotatedasHome-domain-likeHD-ZIPfamilywiththefunctionofpromotingcottonfiberelongationandinitiation.
HAU1355wasassociatedwithFUHML,FSandSFonChr18;itshomologousgenesinG.
arboreumandArabi-dopsisthalianawereCotton_A_16285andAT4G32551.
2,respectively,whichwereannotatedasWD40repeat-like-containingdomainfamilywiththefunctionofpromotingfiberepidermalcellinitiation.
MON-CGR5167wasassoci-atedwithFUHMLandFSonChr11;itshomologousgenesinG.
arboreumandArabidopsisthalianawereCotton_A_07705andAT4G00050.
1,respectively,whichwereannotatedasbasichelix-loop-helix(bHLH)DNA-bindingsuperfamilyproteinfamilytopromotefiberepidermalcellinitiation.
Fiveallelicvariationlociwererelatedtothegenefunc-tionalannotationoffiberqualitytraitsinG.
raimondii(Additionalfile12).
ThehomologousgenesofBNL3436,associatedwithFSonCh25inG.
raimondiiandArabidopsisthaliana,wereCotton_A_07705andAT4G00050.
1,respectively,withthegeneannotationofUDP-glycosy-transferase73B4,whichisinvolvedincellwallsynthesisandfiberdevelopment.
HAU0211andHAU1355werediscoveredwiththesamegenefunctionalannotationasinG.
arboreum.
NAU2564wasassociatedwithFUHML,SFandFSonChr07withthesamegeneasMON_CGR5167inG.
arboreum.
ThehomologousgenesofNAU6627,associatedwithFSonChr21,inG.
raimondiiandArabidopsisthalianawereGorai.
007G150900andAT5G43900.
1,respectively,whichhavegenefunctionrelatedtoMyosin2andmaybeconnectedwiththecellskeleton.
FourallelicvariationlociwererelatedtothegenefunctionalannotationoffiberqualitytraitsinG.
hirsu-tum(Additionalfile13).
HAU1355,MON_CGR5167andHAU0211hadthesamegeneannotation,asde-scribedabove.
STV106,withthehomologousgenesGh_A06G0097andAT1G65910.
1inG.
hirsutumandArabidopsisthaliana,respectively,wasanewlydiscov-eredallelicvariationlociassociatedwithFUHML,MV,FU,SFandFEonChr06,whosegeneannotationwasanNACdomaincontainingprotein28,whichthickensthesecondarywallinArabidopsisthalianaandthexylemandcellwallincotton.
DiscoveryofsuperiorallelesandtypicalmaterialsAccordingtothegenotypedataofthelociassociatedwithfiberquality-relatedtraitsidentifiedatP<0.
05andthephenotypedataoftheBLUPresultsof6fiberquality-relatedtraitsin8environments,48materialswithsuperioralleleswerediscovered(Additionalfile14).
TakingFUHMLasanexample,15markerlociofposi-tivephenotypiceffectsand10markerlociofnegativephenotypiceffectswerefound,withBLUPvaluesran-gingfrom30.
0to31.
48mmandfrom22.
87to25.
98mm,respectively.
NAU1982awastheallelicvari-ationlocuswiththemaximumpositivephenotypiceffect(+0.
473mm)inZY495;meanwhile,MON-CGR6378cwastheallelicvariationlocuswiththemaximumnega-tivephenotypiceffect(1.
23mm)inZY83.
DiscussionPopulationconstructionThepopulationpanelconsistedof503cultivarsinclud-ingsomebasicgermplasmsintroducedfromabroadandevolvedthroughthreevarietyreplacementstages(King,TriceandLxmestarwereintroducedtotheNorthernCottonRegionsin1920sandpartiallyreplacedG.
arbor-eumvarieties;Stoneville4,Delfos531andDPL14re-placedhalfoftheG.
arboreumvarietiesinthe1930sand1940s;DPL15,Stoneville2BandStoneville5AreplacedG.
arboreumvarieties,whichwereplantedforalongtime,andoutdatedG.
hirsutumvarietiesinthe1950s)[26],andbreedingvarietiesfrom1911to2011inChina.
Comparedwiththesamplesizeofthepopulationinpre-viousresearches[24,27],ourpopulationwasmorecom-prehensivethanothersandwaslargerthan500,whichwassufficientforstatisticalpower[22].
ThepopulationpanelincludedcultivarsfromfivemainrepresentativeNieetal.
BMCGenomics(2016)17:352Page8of16cultivatedcottonregionsandwasthusenrichedwithabundantvariationsinyield,fiberqualityanddiseaseresistance.
Inthisstudy,theevaluationofsixfiberquality-relatedtraitsineightenvironmentsshowedwidevariations(0.
80~9.
31%),stableheritability(0.
84~0.
93)(Additionalfile7),andstablechangingtrendsofeachtraitindifferentenvironments(Fig.
4).
PhenotypictraitsanalysisbasedontheBLUPresultsruledoutenviron-mentaleffectsandimprovedtheaccuracyofthecom-plexquantitativetraits.
Boththecompositionofthepopulationandthetraitevaluationindicatedthatthispopulationpanelcouldbeconsideredasanidealre-sourceforassociationmappingofquantitativetraitsinG.
hirsutum.
MoleculargeneticdiversityInthisstudy,atotalof179SSRmarkersproduced426allelelociafterpolymorphismdetectionfrom494genome-wideSSRsinthepopulation.
Theaveragegen-eticdiversityandPICwere0.
377(rangingfrom0.
012to0.
893)and0.
336(rangingfrom0.
012to0.
887)respect-ively,whichwerehigherthanthatofQinetal.
[24].
Ourresearchresultsindicatedthattheselectedmarkershadsufficientpolymorphicinformationtorevealthegeneticrelationshipbetweentheseuplandcottonin-bredcultivars.
Theresultsalsoindicatedthattheaveragegeneticsimilaritycoefficientvariationwas0.
552(rangingfrom0.
337to0.
921),whichcouldbenefitcottonbreedersinselectingparentsinhybridizationbreedingtocreatenovelvariationsandtodevelopnewcultivars.
Compari-sonwithpreviouslypublishedstudies[5–9,28]showedthatthemoleculardiversityinourpopulationwashigher.
Infact,thepreviousresearchersemphasizedthenarrowgeneticbasisofuplandcottonandselectedonlyuplandcottongermplasmsrepresentingpartofthecot-tonplantedareasorbreedingperiods.
Forexample,thestudiesonuplandcottongermplasmsbyAietal.
[9]andNieetal.
[28]onlystudiedtheNorthwestInlandcottonvarieties(inthenorthandsouthofXinjiang),whichweremainlyderivedfromtheformerSovietUnion.
PopulationstructureTheevaluationofthepopulationgeneticstructureisaprerequisiteofgenome-wideassociationstudies[22],becausefalseassociationsusuallycausedbypopulationstructure[29].
Areasonablyaccuratepopulationstruc-turecanleadtomoregeneticsimilaritywithineachgroupandhighergeneticdifferencesamonggroupsandlargelyreducethedefectivesinassociationanalysis[30].
Thus,theaccuracyoftheassociationanalysisdependedonwhetherthepopulationstructurewasappropriate[31,32].
Previously,Abdurakhmonovetal.
usedQ-matrixtoestimateaccessionclusters[23].
Qinetal.
[24]andCaietal.
[33]determinedthepopulationstructureusingSTRUCTUREsoftwaretoshowKvalues.
Wangetal.
[34]determinedthegeneticstructureof55accessionsofG.
barbadensebyseparatingPCAplotsbasedonSSRandSRAPmarkers.
Inthisstudy,thethreemethodsofPCA,unrootedandroundedtree,withK=7corre-spondingtotheuppermoststructurallevelinthecon-tactzonemodel,werecombinedtopredictthepopulationstructurebasedon426allelelocigeneratedby179SSRmarkersdistributingonthe26chromo-somes.
Thepopulationwasclassifiedinto7subpopula-tions,whichwasreasonabletoeliminatethespuriousassociationeffectsintheassociationanalysis.
LinkagedisequilibriumLD,definedasnonrandomcombinationalallelesatdif-ferentgeneloci[29],isthegeneticunderpinningsofas-sociationanalysis.
Inourstudy,itwasshownthatdifferentnumbersoflineararraysoflinkagedisequilib-riumlocidistributedoneachchromosome,especiallythepair-locinumberofLDonChr01,10,12,15,19,21,and26(Fig.
3),onwhichthereweremorepolymorphiclociassociatedwithfiberqualitytraits.
Forexample,amongthe13polymorphiclocionChr01,8markerlociwererelatedtofiberqualitytraits.
Therefore,linkagedisequilibriumwasthebasisoftheassociationanalysis,inwhicheachpair-lociofLDrepresentedagroupallelicvariation.
ThestrongerLDdegreebetweenthelociinthedifferentallelicvariationswithfunctionaldifferences,themoreQTLassociatedwiththeirphenotypeweredetected.
RecombinationwasthegreatestfactorinfluencingLD.
Generally,LDlevelishighinself-pollinatedcrops,andlowincross-pollinationcropsbecauseitwillhaveahigherrecombinationrateandbreakthelinkagedisequi-libriumbetweengeneloci[35,36].
Cottonisacross-pollinationcropwithahigherrecombinationrate;more-over,asignificantamountofhumanbehavior,includingcross,backcross,openpollination,germplasmintroduc-tionandexchangeduringbreedingindifferentcottonplantingregionsgreatlyincreasetherecombinationrate,whichleadstothelowLDlevelinthecottongenome.
Abdurakhmonovetal.
[25]foundthattheattenuationdistanceofLDwas25cM,10cMand30cMatr2≥0.
1,andthedistancereached5–6cM,1–2cMand6–8cMatr2≥0.
2;Zhangetal.
[37]reportedthattheattenuationdistanceofLDwas3.
4cMandlessthan1cMatr2≥0.
1andr2≥0.
2respectively.
Fangetal.
[38]observedthattheattenuationdistanceofLDoftheaveragechromo-somewasapproximately6.
75cM.
Qinetal.
[24]discoveredthattheattenuationdistanceoflinkagedis-equilibriumdecreaseddramaticallyto0–10cM.
Inthisstudy,theattenuationdistanceofLDdecreaseddramat-icallyto0–5cM,asseenfromtheLDattenuationfigureNieetal.
BMCGenomics(2016)17:352Page9of16(Additionalfile5),whichwouldbeusefultofinemap-pingcandidateQTL.
AssociationmappingQTLforfiberqualitytraitsThe216markerloci(P<0.
05)associatedwithsixfibertraitsinthisstudywerecomparedtootherreportedQTLincotton.
Thirteenmarkerlociidentifiedinourstudycoincidedwithpreviousresearch,ofwhich3markerlociweredetectedwiththesametraits(Table2).
NAU3419[39]onChr02andrelatedtoFSwaslocatedatdifferentpositions.
NAU5480[40],dramaticallysig-nificantforFUHMLandwithmuchhigherexplanationofvariation,wasdetectedonChr11.
BNL3436[33]onChr25,significantforFS,wasdetectedbytwopreviousresearchers.
ThecommonQTLdetectedbydifferentre-searchersindicatedthatthemarkerlociassociatedwithfiberqualitytraitsinourstudywerereliable.
ComparingourQTLreferencesequencestoG.
arboreum[41],G.
raimondii[42]andG.
hirsutum[43]showedthat7QTLwererelatedtothegenefunctionoffiberdevelop-ment(Table3).
Recently,Shanetal.
verifiedthatahomeodomain-likesuperfamilyprotein,GhHOX3,couldcontrolcottonfiberelongation[44],andwasassociatedwiththemarkerlocusforfiberlengthanduniformity[45].
Coincidentally,somegenesannotatedasthehomeodomain-likesuperfamilyinthemarkerlocusHAU0211werefoundforfiberuniformityinourstudy.
AgeneannotatedasUDP-glycosyltranserasewasfoundtoberelatedtoBNL3436.
UDP-glycosyltransferasewasre-portedtobeinvolvedintheregulationofcellwallpectinbiosynthesis[46].
ThesecandidategeneswithintheQTLcouldaffectmaturefiberqualitytraitsbyregulatingfibercelldevelopmentandcellwallbiogenesis.
NAU6627(r2=1.
89,P=0.
0110)andMON-CGR5602(r2=1.
89,P=0.
0366),whichwere6.
7cMingeneticdistance,weresignificantlyassociatedwithFSonChr21.
NAU6627,relatedtoageneannotatedasmyosin2,acytoskeletonrelatedgene,waswithinthemarkerlocusBNL3436forTable2ComparisonofQTLassociatedwithfiberqualitytraitstothereportedstudiesLocusOurresearchOtherresearchPosition(cM)P-FDRr2TraitTraitCIR38138.
8684.
81E-020.
89FSLI[61];FE,FL,FS,MIC,FE,MAT,RD,FB[33];FL,MIC,MAT,FR,FB[62];PER,WF,SL[63]NAU3419156.
0344.
20E-021.
00FSFL,FS,FM[38];PB[64]NAU2836156.
2283.
40E-021.
08FUHMLFM[65]8.
69E-043.
17FSNAU5129140.
3223.
92E-021.
06FEBS[66]BNL379073.
21.
59E-021.
53FSSL[67]NAU5480173.
9787.
33E-043.
11FUHMLFS[68];FL,FS[40]4.
85E-021.
12FUNAU267140.
9953.
83E-021.
01FUHMLSY,LY,BPP[69]2.
42E-021.
32FS3.
29E-021.
48SF3.
87E-021.
26FUNAU646828.
4874.
93E-020.
88FUHMLFBN[61]0.
039801.
11MVHAU011902.
41E-021.
30FSFM[65]3.
46E-021.
39SFBNL334743.
2082.
34E-021.
34FSHP[67];MIC,FU,SCI[70]NAU3092236.
3224.
71E-020.
59FUHMLLI[64]0.
037690.
77MVGh27797.
1563.
87E-021.
4SFBNL343603.
45E-020.
78FSFL,FM[71]NAU308423.
9012.
68E-021.
39FUHMLFB[66];BW,NB[64]1.
59E-021.
51FS7.
34E-031.
93FU5.
12E-065.
08SFNieetal.
BMCGenomics(2016)17:352Page10of16fiberstrength.
Therewasareportthatanothercellskel-etoninteractionprotein,GhWLIM1a,couldpromoterfibercellelongationandregulatesfibersecondarycellwallbiogenesis[47].
OverexpressionoftheGhWIM1agenealsocouldincreasematurefiberstrength,whichindicatedthatmyosin2wasanimportantcandidategeneidentifiedfromQTLassociatedwithfiberstrength.
STV106wasre-latedtoanNACdomaincontainingprotein[48],whichcouldregulatesecondarywallbiogenesisandmaypartiallyexplainwhySTV106wasassociatedwithfivemainfiberqualitytraits.
CottonfibercellinitiationwassimilartotrichomeinArabidopsis,whichisregulatedbyWD40,bHLHandMYBtranscriptfactors[49].
WefoundthatsomegenesthatwereannotatedasWD40orbHLHinHAU1355,MONCGR5167andNAV2564.
Therelation-shipsbetweenfiberdevelopment-relatedgenesandma-turefiberqualitytraitsrequirefurtherresearch.
ApplicationsinbreedingTheinformationofthisstudyprovidedthecharacteris-ticsofphenotypicvariationoffiberqualitytraits,thegeneticdiversity,populationstructureandeliteallelesandencouragedustotakeafurtherstudytoproposeadetailedschemeforapplyingtheaboveresultsincottonbreeding.
Firstly,specificandeliteallelesconsideredasselectiontagsofgeneticfragmentsofintrogressionlineswereusedtocharacterizedifferentforegroundselectionpar-entsandwerethencrossedwithininfiltrationlinestoas-sembleeliteallelesfromdifferentmarkerlociintoonerecipientparentandbreedexcellenthybridprogenybyallelicbandsassistedselection.
Forexample,inourstudy,NAU5480(r2=3.
11),MON_DPL0544b(r2=1.
84)andHAU1166(r2=1.
82)associatedwithFUHML,NAU2836(r2=3.
17)andNAU2317(r2=1.
64)associatedwithFS,andDPL0457(r2=1.
46)andNAU6468(r2=1.
11)associatedwithMV.
Theseeliteallelescouldbecombinedbymarker-assistedselectiontodevelopacces-sionswithsuperfiberquality.
Secondly,elitegermplasmscouldbeselectedaspar-entsinthebreedingprogrambasedontheassociationresults;thetypicalcarriermaterialsaggregatedtheallelicvariationwiththemostpositiveefficiency,moredistantgeneticrelationshipandcomplementaryeliteandstablegenetictraits.
Forexample,accordingtothephenotypiceffectsofeachgermplasmbasedontheBLUPofsixfiberqualitytraitsineightenvironments(Additionalfile6),ZY495,ZY488andZY64hadhigherphenotypiceffectsforFUHML,FSandMV,respectively;meanwhile,ZY43hadhigherphenotypiceffectsforFUHMLandrelativelylowerphenotypiceffectsforFS.
ComparingthegeneticsimilaritycoefficientsofZY495,ZY488andZY64toZY43,thelowestgeneticsimilaritycoefficientbetweenZY64andZY43indicatedthattheycouldbeselectedtoimprovefiberquality.
ConclusionsAnassociationmappingpopulationof503nation-wideuplandcottonsinChinawasgenotypedby494genome-wideSSRsandphenotypedineightenviron-ments,whichrevealedabundantmoleculardiversityandphenotypicvariations.
ThepopulationwasdividedTable3AllelicvariationlociassociatedwithfiberqualityandannotatedgenesAllelicvariationlociChromosomeTraitPosition(cM)P-FDRr2GeneannotationHAU021112FU10.
1923.
77E-021.
24Homeodomain-likesuperfamilyproteinSF4.
48E-021.
14HAU135518FUHML118.
0423.
26E-020.
76WD40/YVTNrepeat-like-containingdomainFS1.
72E-021.
08SF3.
40E-020.
96MON_CGR516711FUHML61.
2063.
88E-021.
04basichelix-loop-helix(bHLH),DNA-bindingsuperfamilyproteinBNL343625FS03.
45E-020.
78UDP-glycosyltransferase73B4NAU25647FUHML18.
2233.
05E-021.
17basichelix-loop-helix(bHLH),DNA-bindingsuperfamilyproteinSF2.
71E-021.
63FS3.
53E-032.
33NAU662721FS90.
4341.
10E-021.
89myosin2STV1066FUHML01.
88E-021.
59NACdomaincontainingprotein28MV0.
037751.
19FU4.
59E-021.
17SF2.
77E-021.
75FE4.
38E-021.
06Nieetal.
BMCGenomics(2016)17:352Page11of16intosevensubpopulationsbycomprehensiveanalysis;andtheattenuationdistanceofLDinthispopulationwas0–5cM.
MLMbasedassociationmappingforfiberqualitydetectedknownandnovelelitealleles,andtypicalmaterialswereidentified.
Theresultsinthisstudywillprovideaplatformforfuturegeneticsandbreedingincotton.
MethodsPlantmaterialsAsetof503uplandcottoninbredcultivars(Additionalfile3)constructedthemappingpopulationpanel,whichrepresentsdiversegeneticresourcesrelatedtofiberqual-itytraits.
TheseinbredcultivarsweremainlycollectedfromChinaandtwoforeigncountriesandweredividedintosevencottongrowingregionsaccordingtodifferentecologicalcharacteristics,including79fromtheNNIR,28fromtheNSEMR,225fromtheYRR,141fromtheYtRR,4fromtheSC,20fromtheUnitedStates,and6theSovietUnion.
TheseinbredcultivarsarefrontierspecimensforcultivardevelopmentinChina.
Thecultivarsinthepopulationpanelcoveredfourvarietyimprovementstages:thepreliminaryphaseoftheG.
arboreumcottonanduplandcottonimprovement(1920s–1950s);,thebreedingstageofincreasingyield(late1950s–1970s);thebreedingstageofcomprehen-siveimprovementofyield,fiberqualityanddiseaseresistance(1980s–1990s);andthebreedingstageofgoodfiber,geneticallymodifiedandmechanizedhar-vestingcottonvarieties(1990s–2010s).
The503uplandcottoncultivarswerecollectedfromtheInstituteofCottonResearchofChineseAcademyofAgriculturalSciences(Anyang,China),XinjiangAcademyofAgricultureandReclamationScience(Shihezi,China),ShiheziAcademyofAgriculturalSciences(Shihezi,China),HenanAcademyofAgriculturalSciences(Zhengzhou,China),JiangsuAcademyofAgriculturalSciences(Nanjing,China),CottonResearchInstituteofShanxiAacademyofAgriculturalSciences(Yuncheng,China),ShandongCottonResearchCenter(Jinan,China),andHebeiAgriculturalUniversity(Baoding,China).
Allthecultivarswereauthorizedforonlyscientificresearchpurpose,andweredepositedintheoriginalinstitutesandHuazhongAgriculturalUniversity.
FieldexperimentsandphenotypedatacollectionFieldexperimentswereperformedinShihezi(SHZ),NorthXinjiang(E85.
94°,N44.
27°,theNIR),Kuerle(KEL),SouthXinjiang(E86.
06°,N41.
68°,theNIR),Yuanyang(YY),Henan(E113.
97°,N35.
05°,theYRR),andHuanggang(HG),Hubei(E114.
87°,N30.
44°,theYtRR)in2012and2013.
Theselocationsareincludedinthemaincultivatedcottonareascoveringthecot-tonecologicalareas.
Theuplandcottoncultivarswecollectedarenotendangeredorprotectedspecies.
Thefieldworksandsamplingwereauthorizedbylocalgovernments,respectively.
Standardfieldplotsandagronomictechnologieswereusedtogrowthepopulationsineightenviron-ments(intwoyearsandfourlocations);andfieldtri-alsfollowedacompletelyrandomizedblockdesignwithtworeplicatesineachenvironment.
Takingintoconsiderationofthedifferentcultivationmanagementpatternsinfourlocations,thedesignsofthefieldex-perimentswereconductedasfollows:usingaspecificwide-narrowdistanceplantingpattern,therowspa-cingwas(40+50+46)cm,with9.
5cmbetweenindi-vidualsand50individualsper5minonerow,plantedbysowinginaholeanddrippingirrigationinXinjiang(bothinShiheziandKuerle);therowspa-cingwas90cm,with30cmbetweenindividualsand12individualsper5minonerow,plantedbysowinginholeandfloodinginYuanyang;therowspacingwas100cm,with40cmbetweenindividualsand10individualsper5minonerow,seedlingtransplant-ationandwithoutirrigatinginHuanggang.
Tomeasurethefiberqualitytraits,20bollswerecollectedfromthemiddlefruitbranchesofplantsineachline.
Aftergrinning,10–15gfibersofeachsam-pleweresenttotheInstituteofCottonResearch,ShiheziAcademyofAgriculturalSciencestotestthefiberquality.
Thefiberquality-relatedtraitsweretestedat20°C,and65%relativehumiditywithanHVI1000AutomaticFiberDeterminationSystem,in-cludingthefiberupperhalfmeanlength(FUHML),fiberstrength(FS),micronairevalue(MV),fiberuniformity(FU),shortfiber(SF),andfiberelong-ation(FE).
PhenotypedataanalysisThefiberqualitytraitsstatisticswereobtainedfromone-year-one-locationandeightenvironmentsbybestlinearunbiasedprediction(BLUP)[50].
Broad-senseher-itabilitywascalculatedforeachtraitaccordingtothemethoddescribedbyLiuetal.
[51],whichreferstothepercentageofthetotalvarianceofphenotypeofgeneticvariation,orphenotypevarianceofgeneticvariancepercentage.
Analysisofphenotypicchangingtrendsandrelevanceoffiberqualitytraitswereshowninboxplotform,andthecorrelationpicturesbetweentheenvironmentandenvironmentbythedifferentcodeweredrawnusingthe"R"program.
SPSS17.
0software(http://www-01.
ibm.
com/software/analytics/spss/)wasusedtocalculatebasicstatistics,in-cludingextremevalues,meanvalue,standarddeviation,variablecoefficient,andthestatisticsforthecorrelationbetweentraits.
Nieetal.
BMCGenomics(2016)17:352Page12of16SSRmarkersgenotypingandgeneticdiversityAtotalof494SSRmarkerswereselectedatanaverageof10cMfromtheinterspecificgeneticmapconstructedwith5152markersincotton[12]andwereusedtoscreenthegeneticpolymorphismofthepopulation.
TotalDNAwasextractedfromyoungleavesusingthemodifiedCTABmethod[52].
Poly-merasechainreaction(PCR),electrophoresis(6%denaturingpolyacrylamidegelelectrophoresisforSSRand8%non-denaturingpolyacrylamidegelelectrophoresisforSSCP)andsilverstainingwereperformedaccordingtothemethodsdescribedbyLietal.
[52].
SSRandSSCPfragmentswerecodedas"1"forpresent,"0"forabsent,and""formissingdata.
Thenumberofallelesloci,numberofgenotypes,genediversity,andpolymorphisminformationcontent(PIC)ofpolymorphicmarkerswerecalculatedbyPowermar-kersoftware3.
25[53](http://tree.
bio.
ed.
ac.
uk/software/figtree/).
Geneticdifferentiation(PhiPT)wasestimatedusingtheAMOVAandtheFrequencyfunctionofGENALEX6.
2software[54](http://biology.
anu.
edu.
au/GenAlEx/).
ThehierarchicaldiagramandmatricesofgeneticsimilaritywereobtainedfromtheNTSYS-pcversion2.
10estatisticalpackagebasedonJaccard'salgo-rithms[55].
PopulationstructureUnrootedtreeandcircularclusteringfigureweregener-atedonthebasisofthegenefrequency,geneticdistance,andanalysisofneighbor-joining(NJ)treesfromPower-markersoftware3.
25[53].
Atwo-dimensionaldiagramofprincipalcoordinatesanalysis(PCA)wasproducedbasedonageneticdistance(GD)matrixbyGENA-LEX6.
2software[54].
Thepopulationstructurewasanalyzedandevalu-atedforfurtherassociationmappingbySTRUC-TUREversion2.
1[31,32]usinganadmixturemodel;MCMCwassetto50000,burn-inwassettoarunningtimeof100000andtheKvaluewassetto2–10.
EachKvaluewasrepeatedfivetimes.
Tose-lectasuitableK,drawingLnP(D)andKvalues-changingtrenddiagramswereusedtodeterminewhichKvaluereachedthelargestplateauoftheLnP(D)values.
Iftherewasnosuchplateau,thenthedrawingdelta(K)andKvalues-changingtrenddia-gramswereusedtojudgewhichinflectionpointoftheKvaluewasthemostsuitableforthenumberofthepopulation[56].
Clusteranalysis,principalcom-ponentanalysis(PCA)andstructureanalysiswereusedtodetermineareasonablepopulationstructureandcorrectfalsepositivescausedbythepopulationstructureforassociationmappinginthisstudy.
LinkagedisequilibriumandLDdecayThedistributiondiagramsofthegenome-widelinkagedisequilibrium(LD)attenuationweredrawninSPSSsta-tistics17.
0software(http://www-01.
ibm.
com/software/analytics/spss/)basedonr2andD'fromtheresultsoftheLDrunninginTASSEL[57]andthegeneticdis-tance.
TheLDdecayscatterplotshowstherelationshipbetweenr2andD'onthey-axisandthegeneticdistancebetweenmarkerpairsonthex-axistounderstandthepatternofLDinthevariousannotationgroupsorentiregenome[58].
AssociationanalysisTheexistenceofthepopulationstructureandrelativekinshipinthepopulationalwaysresultinahighlevelofspuriouspositivesinassociationmapping.
Inthisstudy,theTASSELV2.
1[57]softwarepackagewasusedtoanalyzethemolecularmarkerdataandthephenotypicdataofthefiberqualityineightdifferentenvironmentsindividuallyusingthethreemodelsofGLM(P+G+QandP+G+PCA)andMLM(P+G+Q+K).
TheQmatrixwasderivedfromClumppsoftware[59]tomergefiverepetitiveresultsuptotheKvalueforsevenusingtheStructuresoftwarepackage.
KingshipwasgeneratedfromtheTASSELsoftwarepackageresults.
ThePvaluesofmarkersassociatedwithQTLwereregulatedbythemethodofmultipletestingcorrectionbycontrollingthefalsediscoveryrate[60].
Elitealleleexplorationandgene-relatedfunctionalannotationAccordingtotheresultsofthegenome-wideanalysisstudy,theprimersequencesoftheQTLthatweremappedonasinglechromosome,repeatedinmul-tipleenvironmentsandreportedthesameaspreviousstudieswereconsideredtocomparewiththeG.
arboreum[41],G.
raimondii[42]andG.
hirsutum[43]genomesequences,andtheninternalgenespacerandgeneannotationincludingthefunctionandpath-waywerediscovered.
Ethics(andconsenttoparticipate)Notapplicable.
ConsenttopublishNotapplicable.
AvailabilityofdataandmaterialsAllrelevantdataareavailablewithinthemanuscriptanditsadditionalfiles.
Nieetal.
BMCGenomics(2016)17:352Page13of16AdditionalfilesAdditionalfile1:TableS1.
TheSSRmarkersinformationof503cultivars.
(XLSX22kb)Additionalfile2:FigureS1.
Clusteringanalysistreeof503G.
hirsutumcultivarsbasedonJaccard'ssimilaritycoefficientsbySSRs.
(PDF27kb)Additionalfile3:TableS2.
Thepedigreesorigininformationof503cultivars.
(XLS124kb)Additionalfile4:FigureS2.
Neighbor-joiningtreesof503G.
hirsutumcultivars.
Dendrogramof503G.
hirsutumcultivarsbyNJanalysisintheformofunrootedtree.
(a):BasedonNei'sgeneticdistance,thepopulationformed7distinctgroupsfromItoVII,including106,19,147,67,9,103,and52cultivars,respectively;(b):Dendrogramof503G.
hirsutumcultivarsbyNJanalysisintheformofroundtree.
Colorsinthedendrogramcorrespondtothepopulationfromsevendifferentcottonregions.
(TIF5688kb)Additionalfile5:FigureS3.
LDdecaysasafunctionofgeneticdistanceintheassociationpanelconsistingof503uplandcottoncultivars.
(a):LDdecayswithindistanceintheformofr2basedonlinkagedisequilibriumlocionthesamechromosome;(b):LDdecayswithdistanceintheformofr2basedondifferentintervalsoflinkagedisequilibriumlocionthesamechromosome;(c):LDdecayswithdistanceintheformofD'basedonlinkagedisequilibriumlocionthesamechromosome.
(TIF2615kb)Additionalfile6:TableS3.
TheBLUPofsixfiberqualitytraitsof503cultivars.
(XLS243kb)Additionalfile7:TableS4.
Phenotypestatisticsofsixfiberqualitytraitsinthetestedcottoncultivars.
(XLSX12kb)Additionalfile8:TableS5.
Thecorrelationanalysisofsixfiberqualitytraits.
(XLSX10kb)Additionalfile9:FigureS4.
Quantile–quantileplotsofestimated–log10(P-value)fromassociationanalysisusingthreemodels.
(a)FUHML;(b)FS;(c)MV;(d)FU;(e)SF;(f)FE.
Theblacklineistheexpectedlineunderthenulldistribution.
Undertheassumptionthattherearefewtruemarkerassociations,theobservedPvaluesareexpectedtonearlyfollowtheexpectedPvalues.
(TIF8527kb)Additionalfile10:TableS6.
Themarkerlocirelatedtosixfiberqualitytraitsofphenotypicvariation.
(XLSX18kb)Additionalfile11:TableS7.
TheresultsofcomparisonoftheallelicvariationlocirelatedfiberqualityandtheGossypiumarboreumgenome.
(XLSX14kb)Additionalfile12:TableS8.
Theresultsofthecomparisonoftheallelicvariationloci-relatedfiberqualityandtheGossypiumraimondiigenome.
(XLSX14kb)Additionalfile13:TableS9.
Theresultsofthecomparisonoftheallelicvariationloci-relatedfiberqualityandtheGossypiumhirsutumgenome.
(XLSX14kb)Additionalfile14:TableS10.
TheSSRmarkerlociassociatedwithsixfiberqualitytraitsofphenotypiceffectsandtypicalmaterials.
(XLSX15kb)AbbreviationsBLUP:bestlinearunbiasedprediction;FE:fiberelongation;FS:fiberstrength;FU:fiberuniformity;FUHML:fiberupperhalfmeanlength;LD:linkagedisequilibrium;MLM:mixedlinearmodel;MV:micronairevalue;NIR:NorthwesternInlandRegion;NSEMR:NorthernSpecificEarlyMaturationRegion;QTL:quantitativetraitloci;SCR:SouthChinaRegion;SF:shortfiber;SSR:simplesequencerepeat;YRR:YellowRiverRegion;YtRR:YangtzeRiverRegion.
CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.
Authors'contributionXHNconductedtheexperimentswithCH;WXZandCSconductedassociationanalysisandco-wrotethemanuscripttext.
ZXLdesignedandsupervisedtheresearchwithXLZandrevisedthemanuscript.
XHN,CYY,WL,CH,WXZ,CS,BBZ,HTW,ZHYandBSDinvestigatedphenotypictraitsinKuerle,Shihezi,YuyangandHuanggang(in2012and2013,respectively),andconducteddataanalysis.
MJWanalyzedtheQTLrelatedgenesinthecottongenome.
Allauthorsreviewedthemanuscript.
Alltheauthorshadreadandapprovedthefinalversionofthemanuscript.
Authors'informationXHN,CH,WXZ,CS,BBZ,HTW,ZHY,BSD,MJW,XLZ,andZXL:NationalKeyLaboratoryofCropGeneticImprovement,CollegeofPlantScience&Technology,HuazhongAgriculturalUniversity,Wuhan430070,Hubei,China;XHN,andCYY:CottonResearchInstitute,ShiheziAcademyofAgricultureScience,Shihezi832011,Xinjiang,China.
WL:EconomicCropResearchInstitute,HenanAcademyofAgriculturalSciences,Zhengzhou450002,Henan,China.
AcknowledgementsWewouldliketothanktheInstituteofCottonResearchofCAAS(Anyang,Henan),YuYu(CottonInstitute,XinjiangAcademyofAgricultureandReclamationScience,Shihezi,Xinjiang),ChunyuanYou(CottonInstitute,ShiheziAcademyofAgriculturalSciences,Xinjiang),WuLi(HenanAcademyofAgriculturalSciences,Zhengzhou,Henan),XinlianShen(EconomicCropsResearchInstitute,JiangsuAcademyofAgriculturalSciences,Nanjing,Jiangsu),YanliJiang(CottonResearchInstituteofShanxiAacademyofAgriculturalSciences,Yuncheng,Shanxi),ZhiweiWang(ShandongCottonResearchCenter,Jinan,Shandong),andZhikunLi(HebeiAgriculturalUniversity,Baoding,Hebei)whoprovideduscottoncultivarseeds.
WealsowouldliketothankYuYu,JilianLi,XinzhuNing,JunjiSu(CottonInstitute,XinjiangAcademyofAgricultureandReclamationScience,Shihezi,Xinjiang),JianghongQin(CottonInstitute,ShiheziAcademyofAgriculturalSciences,Xinjiang),andWuLiandDeyiXie(HenanAcademyofAgriculturalSciences,Zhengzhou,Henan)whoprovidedusexperimentalfieldsandinvestigatedthephenotypictraitsinKuerle,ShiheziandYuyang.
FundingThisworkwasfinanciallysupportedbytheFundamentalResearchFundsfortheCentralUniversities(GrantNo.
2014PY015).
Authordetails1NationalKeyLaboratoryofCropGeneticImprovement,CollegeofPlantSciences&Technology,HuazhongAgriculturalUniversity,Wuhan430070,Hubei,China.
2CottonResearchInstitute,ShiheziAcademyofAgricultureScience,Shihezi832011,Xinjiang,China.
3EconomicCropResearchInstitute,HenanAcademyofAgriculturalSciences,Zhengzhou450002,Henan,China.
Received:9November2015Accepted:23April2016References1.
ChenZJ,SchefflerBE,DennisE,TriplettBA,ZhangT,GuoW,etal.
Towardsequencingcotton(Gossypium)genomes.
PlantPhysiol.
2007;145:1303–10.
2.
JiDP.
CottonEncyclopedicKnowledge.
Beijing:ChinaAgriculturalPress;2001.
3.
ZhouSH.
GenogramofcottonvarietiesinChina.
Sichuan:SichuanScienceandTechnologyPress;2000.
4.
HuangZK.
CottonvarietiesandtheirgenealogyinChina.
Beijing:ChineseAgriculturalPress;2007.
5.
XuQH,ZhangXL,NieYC.
StudyoncomparisonofgeneticdiversitybetweentheYangtzeRiverandYellowRivervalley.
ActaGenetSin.
2001;28:683–90.
6.
WenXJ,MaCY,WangSF,ZhangGY,LiXH.
MoleculardivergenceestimationofChineseuplandcottonswithfusariumandverticilliumwiltResistance.
ActaAgronSin.
2005;38:936–43.
7.
WuDP,FangXX,CuiRG,ChenJH,ZhuSJ.
GeneticrelationshipanddiversityoftheuplandcottongermplasmsfromdifferentcottonproducingcountriesusingSSRmarkers.
CottonSci.
2011;23:291–9.
8.
ZhaoZS,YuSX,FanSL,ZhangWG,PangCJ,MeiYJ.
Analysisofgeneticdiversityofearlymaturinguplandcottonvarietiesinnorthern.
CottonSci.
2012;24:473–80.
Nieetal.
BMCGenomics(2016)17:352Page14of169.
AiXT,LiangYJ,ShaH,WangJZ,ZhengJY,TuEXJ,etal.
GeneticdiversityanalysisonlocaluplandcottoncultivarsinXinjiangbasedonSSRmarkers.
ActaAgronSin.
2014;40:369–79.
10.
KongFL.
Quantitativegeneticsinplants.
Beijing:ChinaAgriculturalUniversityPress;2005.
11.
YuY,YuanDJ,LiangSG,LiXM,WangXQ,LinZX,etal.
Genomestructureofcottonrevealedbyagenome-wideSSRgeneticmapconstructedfromaBC1populationbetweenGossypiumhirsutumandG.
barbadense.
BMCGenomics.
2011;12:15.
12.
LiXM.
Constructionofintrogressionlines,developmentofnewmarkers,andconstructionofhigh-densitygeneticlinkagemapincotton.
HuazhongAgriculturalUniversitylibrary.
2013.
13.
LiuRZ,AiNJ,ZhuXX,LiuFJ,GuoWZ,ZhangTZ.
Geneticanalysisofplantheightusingtwoimmortalizedpopulationsof"CRI12*J8891"inGossypiumhirsutumL.
Euphytica.
2014;196:51–61.
14.
NingZY,ChenH,MeiHX,ZhangTZ.
MoleculartaggingofQTLsforfiberqualityandyieldintheuplandcottoncultivarAcala-Prema.
Euphytica.
2014;195:143–56.
15.
FangDD,JenkinsJN,DengDD,McCartyJC,LiP,WuJX.
Quantitativetraitlocianalysisoffiberqualitytraitsusingarandom-matedrecombinantinbredpopulationinUplandcotton(GossypiumhirsutumL.
).
BMCGenomics.
2014;15:397.
16.
AiNJ,ZhuXX,GuanRZ,ZhaoJJ,ZhangTZ.
Geneticanalysisofmajorlocusgroupconstitutionsofgrowthstagesinuplandcotton.
SciAgricSin.
2010;43:4140–8.
17.
BolekY,El-ZikKM,PepperAE,BellAA,MagillCW,ThaxtonPM,etal.
Mappingofverticilliumwiltresistancegenesincotton.
PlantSci.
2005;168:1581–90.
18.
WangHM,LinZX,ZhangXL,ChenW,GuoXP,NieYC,etal.
Mappingandquantitativetraitlocianalysisofverticilliumwiltresistancegenesincotton.
JIntegrPlantBiol.
2008;50:174–82.
19.
JiangF,ZhaoJ,ZhouL,GuoWZ,ZhangTZ.
MolecularmappingofverticilliumwiltresistanceQTLclusteredonchromosomesD7andD9inuplandcotton.
SCIChinaSerC.
2009;52:872–84.
20.
FangH,ZhouHP,SanogoS,FlynnR,PercyRG,HughsSE,etal.
Quantitativetraitlocusmappingforverticilliumwiltresistanceinabackcrossinbredlinepopulationofcotton(Gossypiumhirsutum*Gossypiumbarbadense)basedonRGA-AFLPanalysis.
Euphytica.
2013;194:79–91.
21.
ZhaoYL,WangHM,ChenW,LiYH.
Geneticstructure,linkagedisequilibriumandassociationmappingofverticilliumwiltresistanceinelitecotton(GossypiumhirsutumL.
)germplasmpopulation.
PlosOne.
2014;9:e86308.
22.
YangXH,YanJB,ZhengYP,YuJM,LiJS.
Reviewsofassociationanalysisforquantitativetraitsinplants.
ActaAgronSin.
2007;33(4):523–30.
23.
AbdurakhmonovIY,SahaS,JenkinsJN,BurievZT,ShermatovSE,SchefflerBE,etal.
LinkagedisequilibriumbasedassociationmappingoffiberqualitytraitsinG.
hirsutumL.
varietygermplasm.
Genetica.
2009;136:401–17.
24.
QinHD,ChenM,YiXD,BieS,ZhangC,ZhangYC,etal.
IdentificationofassociatedSSRmarkersforyieldcomponentandfiberqualitytraitsbasedonframemapanduplandcottoncollections.
PlosOne.
2015;10:e0118073.
25.
AbdurakhmonovIY,KohelRJ,YuJZ,PepperAE,AbdullaevAA,KushanovFN,etal.
MoleculardiversityandassociationmappingoffiberqualitytraitsinexoticG.
hirsutumL.
germplasm.
Genomics.
2008;92:478–87.
26.
YuSX.
GeneticsandBreedingofCottoninChina.
Shandong:ShandongScienceandTechnologyPress;2003.
27.
NieXH,YouCY,BaoJ,LiXF,HuiH,LiuHL,etal.
ExplorationofeliteallelesofagronomicandfiberqualitytraitsinXinluzaocottonvarietiesbyassociationanalysis.
SciAgricSin.
2015;48(15):2891–910.
28.
NieXH,YouCY,LiXF,QinJH,HuangC,GuoHL,etal.
ConstructionofDNAfingerprintingandanalysisofgeneticdiversityforXinluzaocottonvarieties.
ActaAgronSin.
2014;40:2113–26.
29.
Flint-GarciaSA,ThornsberryJM,BucklerES.
Structureoflinkagedisequilibriuminplants.
AnnuRevofPlantBiol.
2003;54:357–74.
30.
WeiSP,LiuXF,YangSX,LuHY,NiuY,ZhangYM.
ComparisonofvariousclusteringmethodsforpopulationstructureinChinesecultivatedsoybean(Glycinemax(L.
)Merr.
).
JNanjingAgriUniv.
2011;34:13–7.
31.
PritchardJK,StephensM,DonnellyP.
Inferenceofpopulationstructureusingmultilocusgenotypedata.
Genetics.
2000;155:945–59.
32.
PritchardJK,StephensM,RosenbergNA,DonnellyP.
Associationmappinginstructuredpopulations.
AmJHumGenet.
2000;67:170–81.
33.
CaiCP,YeWX,ZhangTZ,GuoWZ.
AssociationanalysisoffiberqualitytraitsandexplorationofeliteallelesinUplandcottoncultivars/accessions(GossypiumhirsutumL.
).
JIntegrPlantBiol.
2014;56:51–62.
34.
WangXQ,YuY,LiW,GuoHL,LinZX,ZhangXL.
AssociationanalysisofyieldandfiberqualitytraitsinGossypiumbarbadensewithSSRsandSRAPs.
GenetMolRes.
2013;12(3):3353–62.
35.
GautBS,LongAD.
Thelowdownonlinkagedisequilibrium.
PlantCell.
2003;15:1502–6.
36.
ShifmanS,KuypersJ,KokorisM,YakirB,DarvasiA.
Linkagedisequilibriumpatternsofthehumangenomeacrosspopulations.
HumMolGenet.
2003;12:771–6.
37.
ZhangYC,BieS,Xian-DaYI,ZhangC,Cheng-QiLI,QinHD.
PopulationStructureandLinkagedisequilibriumanalysisofgermplasmresourcesinuplandcotton.
CottonSci.
2011;23:500–6.
38.
FangDD,HinzeLL,PercyRG,LiP,DengD,ThyssenG.
Amicrosatellite-basedgenome-wideanalysisofgeneticdiversityandlinkagedisequilibriuminuplandcotton(GossypiumhirsutumL.
)cultivarsfrommajorcottongrowingcountries.
Euphytica.
2013;191:391–401.
39.
CaoZB,WangP,ZhuXF,ChenH,ZhangTZ.
SSRmarker-assistedimprovementoffiberqualitiesinGossypiumhirsutumusingG.
barbadenseintrogressionlines.
TheorApplGenet.
2014;127:587–94.
40.
YuJW,ZhangK,LiSY,YuSX,ZhaiHH,WuM,etal.
MappingquantitativetraitlociforlintyieldandfiberqualityacrossenvironmentsinaGossypiumhirsutum*Gossypiumbarbadensebackcrossinbredlinepopulation.
TheorApplGenet.
2013;126:275–87.
41.
LiFG,FanGY,WangKB,SunFM,YuanYL,SongGL,etal.
GenomesequenceofthecultivatedcottonGossypiumarboreum.
NatGenet.
2014;46:567–72.
42.
PatersonAH,WendelJF,GundlachH,GuoH,JenkinsJ,JinDC,etal.
RepeatedpolyploidizationofGossypiumgenomesandtheevolutionofspinnablecottonfibres.
Nature.
2012;492:423–7.
43.
ZhangTZ,HuY,JiangWK,FangL,GuanXY,ChenJD,etal.
Sequencingofallotetraploidcotton(GossypiumhirsutumL.
acc.
TM-1)providesaresourceforfiberimprovement.
NatBiotechnol.
2015;33:531–U252.
44.
ShanCM,ShangguanXX,ZhaoB,ZhangXF,ChaoLM,YangCQ,etal.
ControlofcottonfibreelongationbyahomeodomaintranscriptionfactorGhHOX3.
NatCommun.
2014;5:5519.
45.
LuanM,GuoX,ZhangY,YaoJ,ChenW.
QTLmappingforagronomicandfibertraitsusingtwointerspecificchromosomesubstitutionlinesofUplandcotton.
PlantBreed.
2009;128:671–9.
46.
QinLX,RaoY,LiL,HuangJF,XuWL,LiXB.
CottonGalT1encodingaputativeglycosyltransferaseisinvolvedinregulationofcellwallpectinbiosynthesisduringplantdevelopment.
PlosOne.
2013;8:e59115.
47.
HanLB,LiYB,WangHY,WuXM,LiCL,LuoM.
ThedualfunctionsofWLIM1aincellelongationandsecondarywallformationindevelopingcottonfibers.
PlantCell.
2013;25:4421–38.
48.
WangHZ,ZhaoQ,ChenF,WangMY,DixonRA.
NACdomainfunctionandtranscriptionalcontrolofasecondarycellwallmasterswitch.
PlantJ.
2011;68:104–1114.
49.
PeschM,HülskampM.
One,two,threemodelsfortrichomepatterninginArabidopsisCurrOpininPlantBiol.
2009;12:587–92.
50.
TechnowF,RiedelsheimerC,SchragTA,SchragTA,MelchingerAE.
Genomicpredictionofhybridperformanceinmaizewithmodelsincorporatingdominanceandpopulationspecificmarkereffects.
TheorApplGenet.
2012;125:1181–94.
51LiuLF,MaSX,HuangZY.
Thecropquantitativegenetics.
Beijing:Agriculturepress;1984.
52.
LiXM,YuanDJ,WangHT,ChenXM,WangB,LinZX,etal.
IncreasingcottongenomecoveragewithpolymorphicSSRsasrevealedbySSCP.
Genome.
2012;55(6):459–70.
53.
LiuKJ,MuseSV.
PowerMarker:integratedanalysisenvironmentforgeneticmarkerdata.
Bioinformatics.
2005;21:2128–9.
54.
PeakallR,SmousePE.
GENALEX6:geneticanalysisinExcel.
Populationgeneticsoftwareforteachingandresearch.
MolEcolNotes.
2006;6:288–95.
55.
AdamsDC,RohlfFJ.
EcologicalcharacterdisplacementinPlethodon:biomechanicaldifferencesfoundfromageometricmorphometricstudy.
PNatlAcadSciUSA.
2000;97:4106–11.
56.
MezmoukS,DubreuilP,BosioM,DécoussetL,CharcossetA,PraudS,etal.
Effectofpopulationstructurecorrectionsontheresultsofassociationmappingtestsincomplexmaizediversitypanels.
TheorApplGenet.
2011;122:1149–60.
Nieetal.
BMCGenomics(2016)17:352Page15of1657.
BradburyPJ,ZhangZ,KroonDE,CasstevensTM,RamdossY,BucklerES.
TASSEL:softwareforassociationmappingofcomplextraitsindiversesamples.
Bioinformatics.
2007;23:2633–5.
58.
XiaoYJ,CaiDF,YangW,YeW,YounasM,WuJS,etal.
Geneticstructureandlinkagedisequilibriumpatternofarapeseed(BrassicanapusL.
)associationmappingpanelrevealedbymicrosatellites.
TheorApplGenet.
2012;125:437–47.
59.
JakobssonM,RosenbergNA.
CLUMPP:aclustermatchingandpermutationprogramfordealingwithlabelswitchingandmultimodalityinanalysisofpopulationstructure.
Bioinformatics.
2007;23:1801–6.
60.
BenjaminiY,HochbergY.
Controllingthefalsediscoveryrate:apracticalandpowerfulapproachtomultipletesting.
JRStatSocSerCApplStat.
1995;57:289–300.
61.
LiuRZ,WangBH,GuoWZ,QinYS,WangLG,ZhangYM,etal.
QuantitativetraitlocimappingforyieldanditscomponentsbyusingtwoimmortalizedpopulationsofaheterotichybridinGossypiumhirsutumL.
MolBreeding.
2012;29:297–311.
62.
WangBH,GuoWZ,ZhuXF,WuYT,HuangNT,ZhangTZ.
QTLmappingoffiberqualityinanelitehybridderived-RILpopulationofuplandcotton.
Euphytica.
2006;152:367–78.
63.
WuJX,GutierrezOA,JenkinsJN,McCartyJC,ZhuJ.
QuantitativeanalysisandQTLmappingforagronomicandfibertraitsinanRIpopulationofuplandcotton.
Euphytica.
2009;165:231–45.
64.
ZhangTZ,QianN,ZhuXF,ChenH,WangS,MeiHX,etal.
VariationsandtransmissionofQTLallelesforyieldandfiberqualitiesinUplandcottoncultivarsdevelopedinChina.
PlosOne.
2013;8:e57220.
65.
SunFD,ZhangJH,WangSF,GongWK,ShiYZ,LiuAY,etal.
QTLmappingforfiberqualitytraitsacrossmultiplegenerationsandenvironmentsinuplandcotton.
MolBreeding.
2012;30:569–82.
66.
GuoX,GuoYP,MaJ,WangF,SunMZ,GuiLJ,etal.
MappingheteroticlociforyieldandagronomictraitsusingchromosomesegmentintrogressionlinesinCotton.
JIntegrPlantBiol.
2013;55:759–74.
67.
AnCF,JenkinsJN,WuJX,GuoYF,McCartyJC.
UseoffiberandfuzzmutantstodetectQTLforyieldcomponents,seed,andfibertraitsofuplandcotton.
Euphytica.
2010;172:21–34.
68.
ZhangXJ,YuanYC,WeiZ,GuoX,GuoYP,ZhangSQ,etal.
MolecularmappingandvalidationofamajorQTLconferringresistancetoadefoliatingisolateofverticilliumwiltincotton(GossypiumhirsutumL.
).
PlosOne.
2014;9:e96226.
69.
ZhangW,LiuF,LiSH,WangW,WangC,ZhangX,etal.
QTLAnalysisonyieldanditscomponentsinrecombinantinbredlinesofuplandcotton.
ActaAgronSin.
2011;3:433–42.
70.
ShenXL,ZhangTZ,GuoWZ,ZhuXF,ZhangXY.
MappingfiberandyieldQTLswithmain,epistatic,andQTL*environmentinteractioneffectsinrecombinantinbredlinesofuplandcotton.
CropSci.
2006;46:61–6.
71.
ShenXL,GuoWZ,ZhuXF,YuanYL,YuJZ,KohelRJ,etal.
MolecularmappingofQTLsforfiberqualitiesinthreediverselinesinUplandcottonusingSSRmarkers.
MolBreeding.
2005;15:169–18.
Weacceptpre-submissioninquiriesOurselectortoolhelpsyoutondthemostrelevantjournalWeprovideroundtheclockcustomersupportConvenientonlinesubmissionThoroughpeerreviewInclusioninPubMedandallmajorindexingservicesMaximumvisibilityforyourresearchSubmityourmanuscriptatwww.
biomedcentral.
com/submitSubmityournextmanuscripttoBioMedCentralandwewillhelpyouateverystep:Nieetal.
BMCGenomics(2016)17:352Page16of16
Toexplorethemoleculardiversity,populationstructureandelitealleles,503inbredcultivarsdevelopedinChinaandsomeforeigncultivarsfromtheUnitedStatesandtheSovietUnionwerecollectedandanalyzedby494genome-wideSSRs(SimpleSequenceRepeats).
Methods:Fourhundredandninety-fourpairsofSSRswithhighpolymorphismanduniformdistributionon26chromosomeswereusedtoscanpolymorphismsin503nation-wideuplandcottons.
TheprogramminglanguageRwasusedtomakeboxplotsforthephenotypictraitsindifferentenvironments.
Molecularmarkerdataand6fiberqualitytraitswereanalyzedbythemethodofMLM(mixedlinearmodel)(P+G+Q+K)intheTASSELsoftwarepackageonthebasisofthepopulationstructureandlinkagedisequilibriumanalysis.
Thelociofeliteallelicvariationandtypicalmaterialscarryingelitealleleswereidentifiedbasedonphenotypiceffectvalues.
Results:Atotalof179markerswerepolymorphicandgenerated426alleleloci;thepopulationbasedonmoleculardiversitywasclassifiedintosevensubpopulationscorrespondingtopedigreeorigin,ecologicalandgeographicaldistribution.
Theattenuationdistanceoflinkagedisequilibriumdroppedsignificantlyupto0–5cM.
Associationmappingforfiberqualityshowedthat216markerlociwereassociatedwithfiberqualitytraits(PFE(0.
581)>FS(0.
474)>MV(0.
445)>FU(0.
410)>SF(0.
380).
Itwasmoreimportanttofurtheranalyzeonetraitbetweentwoenvironments;takingFig.
1PCAplotsof503uplandcottoncultivarsbasedonSSRs.
Bluetriangles,greensquares,browntriangles,purplediamonds,lightbluesquare,reddiamondsandyellowcirclesrepresentcultivarsfromdifferentregionsFig.
2PopulationstructureanalysisandlineschartoftheKvaluewithLnP(D)valueandΔKvaluebasedonstructureanalysis.
LinechartoftheLnP(D)valuewiththechangeofK(a);LinechartoftheΔKvaluewiththechangeofK(b);Populationstructureof503cultivar-basedSSRmarkers(c).
Q-plotshowingtheclusteringof503uplandcottoncultivarsbasedontheanalysisofgenotypicdatausingSTRUCTURE.
Eachcultivarisrepresentedbyaverticalbar.
Thecoloredsubsectionswithineachverticalbarindicatethemembershipcoefficient(Q)ofthecultivartodifferentclustersNieetal.
BMCGenomics(2016)17:352Page4of16FUHMLasexample(Fig.
5a),thecorrelationsrangedfrom0.
40to0.
76,andthecorrelationwas0.
76forFUHML_12KELandFUHML_13KEL.
Theredlinethatwasnearthe45°asthelineofgreatestslopein-dicatedacorrelationbetweenFUHML_12KELandFUHML_13KEL.
Associationmappingoffiberquality-relatedtraitsBasedonthegenotypedata,thePCAmatrix,thekinshipmatrix,andthefiberqualitytraitsdataoftheBLUPre-sultsin8environments,amixedlinearmodelwasusedtoanalyzethemarker-traitassociations.
Duringassoci-ationmapping,threemodels,GLM(P+G)+Q,GLMFig.
3ThedistributionofLDamong426SSRlocion26chromosomesin503uplandcottoncultivars(R2value).
SSRmarkerswerealongtheX-axis.
Eachpixelabovethediagonalrepresentsther2sizeofthecorrespondingpairsofmarkers,asshowninthecolorcodeattheupperright,andeachpixelbelowthediagonalrepresentstheP-valuesizeoftestingtheLDatthelowerrightFig.
4TheboxplotsofthechangingtrendsofsixfiberqualitytraitsineightenvironmentsNieetal.
BMCGenomics(2016)17:352Page5of16(P+G)+PCs,andMLM(G+P+Q+K),werecom-paredwitheachotherintheassociationanalysis(Additionalfile9).
ThecontroleffectofthepopulationstructureforFUHML,FS,SF,andFEinthethreemodelswassimilar.
However,theMVanalysisusingtheMLM-Q-Kmodelwassuperiortotheothertwomodels,andtheFUanalysisusingtheGLM-QandMLM-Q-KmodelswasbetterthantheofGLM-PCAmodel.
Accordingtotheabovecomparisonresults,theMLM-Q-Kmodelhadbetterperformance.
Atotalof179SSRmarkerswereusedformarker-traitassociationafterfilteringfor5%minimumalleles,amongwhich91(50.
84%)markerswereassociatedwithfiberqualitytraitsattheP<0.
05level.
FifteenmarkersweresignificantlyassociatedattheP<0.
01level(Fig.
6).
Anaverageof3.
5markerswasdetectedoneachchromosome(rangingfrom1to8),withthemaximumof8markersonChr01andChr19.
Onemarkerwasgenerallyassociatedwithseveraltraits.
Forexample,MON_DC40013onChr01wasrelatedtoFUHML,SF,FUandFS,andNAU2564onChr07wasrelatedtoFUHML,SFandFS.
Therewere216lociassociatedwithfiberqualitycom-ponentsattheP<0.
05significancelevel,amongwhich27weresignificantattheP<0.
01level.
Therangeofphenotypicvariationexplanation(PVE)observedwasfrom0.
58%(MON_DPL0042b)to5.
12(NAU3084c),withanaverageof2.
70%(Additionalfile10).
Amongthe6traits,FSwasassociatedwiththemostloci,uptoamaximumof61(P<0.
05)and7(P<0.
01);PVErangedfrom0.
58(MON-DPL0042b,P<0.
05)to3.
17%(NAU2836a,P<0.
001),withameanof2.
63%.
TheremarkablecontributionlociwereNAU2858a(2.
92%),BNL3089a(2.
86%)andMONCGR5399c(2.
73%),especiallyattheP<0.
001level.
FUHMLwasassociatedwiththesecondnumberofloci,uptoamaximumof46(P<0.
05)and4(P<0.
01);PVErangedfrom0.
59(NAU3092a,P<0.
05)to3.
11%(NAU5480b,P<0.
001),withameanof2.
35%,inwhichNAU5480a(2.
75%)hadasignificantcontribu-tionatP<0.
001.
Therewereupto42(P<0.
05)and6(P<0.
01)lociassociatedwithSF,andPVEsrangedfrom0.
68(HAU2835b,P<0.
05)to5.
12%(NAU3084c,P<0.
001),Fig.
5ThecorrelationsofsixfiberqualitytraitsineightenvironmentsNieetal.
BMCGenomics(2016)17:352Page6of16Fig.
6Thedistributionofthelocatedmarkersassociatedwiththefiberquality-relatedtraitson26chromosomesNieetal.
BMCGenomics(2016)17:352Page7of16withameanof3.
80%.
NAU3084b(5.
08%),MON-DPL0893a(4.
43%),andMON-DC40013b(3.
03%)con-tributedprominentlytoSFatP<0.
001.
Therewere25(P<0.
05)and3(P<0.
01)lociassociatedwithFE,withPVErangingfrom0.
70(MON-CGR5423b,P<0.
05)to2.
99%(HAU4806-SSCPa,P<0.
001).
BNL846bandMON_CGR5113bcontributedtoFEatP<0.
01.
Therewere23(onlyatP<0.
05)lociassociatedwithMV,withPVErangingfrom0.
66(NAU3138b)to1.
46%(DPL0457a);NBRI-HQ524733bcontributingtoMVwasdetectedatP<0.
05.
Therewere19(P<0.
05)and7(P<0.
01)lociassociatedwithFU,andthePVEsrangedfrom1.
06(HAU1279d)to2.
51%(HAU1166a),withameanof2.
14%.
ThecontributionlociwereobservedinMON-CGR5602a(2.
39%),MON-DPL0893a(2.
26%),MON-DC40013b(2.
21%),NAU3084c(1.
93%),HAU1166b(1.
87%)andNAU3084b(1.
82%)atP<0.
01.
Exploringeliteallele-relatedgenesinthecottongenomeThereferencesequencesof91eliteallelelociassociatedwithfiberqualitytraitswereexploredbasedonrelatedgenesinG.
arboreum,G.
raimondiiandG.
hirsutum,.
Threeallelicvariationlociwererelatedtogenefunc-tionalannotationoffiberqualitytraitsinG.
arboreum(Additionalfile11).
HAU0211wasassociatedwithFUandSFonChr12;itshomologousgenesinG.
arboreumandArabidopsisthalianawereCotton_A_01461andAT5G16560.
1,respectively,whichwereannotatedasHome-domain-likeHD-ZIPfamilywiththefunctionofpromotingcottonfiberelongationandinitiation.
HAU1355wasassociatedwithFUHML,FSandSFonChr18;itshomologousgenesinG.
arboreumandArabi-dopsisthalianawereCotton_A_16285andAT4G32551.
2,respectively,whichwereannotatedasWD40repeat-like-containingdomainfamilywiththefunctionofpromotingfiberepidermalcellinitiation.
MON-CGR5167wasassoci-atedwithFUHMLandFSonChr11;itshomologousgenesinG.
arboreumandArabidopsisthalianawereCotton_A_07705andAT4G00050.
1,respectively,whichwereannotatedasbasichelix-loop-helix(bHLH)DNA-bindingsuperfamilyproteinfamilytopromotefiberepidermalcellinitiation.
Fiveallelicvariationlociwererelatedtothegenefunc-tionalannotationoffiberqualitytraitsinG.
raimondii(Additionalfile12).
ThehomologousgenesofBNL3436,associatedwithFSonCh25inG.
raimondiiandArabidopsisthaliana,wereCotton_A_07705andAT4G00050.
1,respectively,withthegeneannotationofUDP-glycosy-transferase73B4,whichisinvolvedincellwallsynthesisandfiberdevelopment.
HAU0211andHAU1355werediscoveredwiththesamegenefunctionalannotationasinG.
arboreum.
NAU2564wasassociatedwithFUHML,SFandFSonChr07withthesamegeneasMON_CGR5167inG.
arboreum.
ThehomologousgenesofNAU6627,associatedwithFSonChr21,inG.
raimondiiandArabidopsisthalianawereGorai.
007G150900andAT5G43900.
1,respectively,whichhavegenefunctionrelatedtoMyosin2andmaybeconnectedwiththecellskeleton.
FourallelicvariationlociwererelatedtothegenefunctionalannotationoffiberqualitytraitsinG.
hirsu-tum(Additionalfile13).
HAU1355,MON_CGR5167andHAU0211hadthesamegeneannotation,asde-scribedabove.
STV106,withthehomologousgenesGh_A06G0097andAT1G65910.
1inG.
hirsutumandArabidopsisthaliana,respectively,wasanewlydiscov-eredallelicvariationlociassociatedwithFUHML,MV,FU,SFandFEonChr06,whosegeneannotationwasanNACdomaincontainingprotein28,whichthickensthesecondarywallinArabidopsisthalianaandthexylemandcellwallincotton.
DiscoveryofsuperiorallelesandtypicalmaterialsAccordingtothegenotypedataofthelociassociatedwithfiberquality-relatedtraitsidentifiedatP<0.
05andthephenotypedataoftheBLUPresultsof6fiberquality-relatedtraitsin8environments,48materialswithsuperioralleleswerediscovered(Additionalfile14).
TakingFUHMLasanexample,15markerlociofposi-tivephenotypiceffectsand10markerlociofnegativephenotypiceffectswerefound,withBLUPvaluesran-gingfrom30.
0to31.
48mmandfrom22.
87to25.
98mm,respectively.
NAU1982awastheallelicvari-ationlocuswiththemaximumpositivephenotypiceffect(+0.
473mm)inZY495;meanwhile,MON-CGR6378cwastheallelicvariationlocuswiththemaximumnega-tivephenotypiceffect(1.
23mm)inZY83.
DiscussionPopulationconstructionThepopulationpanelconsistedof503cultivarsinclud-ingsomebasicgermplasmsintroducedfromabroadandevolvedthroughthreevarietyreplacementstages(King,TriceandLxmestarwereintroducedtotheNorthernCottonRegionsin1920sandpartiallyreplacedG.
arbor-eumvarieties;Stoneville4,Delfos531andDPL14re-placedhalfoftheG.
arboreumvarietiesinthe1930sand1940s;DPL15,Stoneville2BandStoneville5AreplacedG.
arboreumvarieties,whichwereplantedforalongtime,andoutdatedG.
hirsutumvarietiesinthe1950s)[26],andbreedingvarietiesfrom1911to2011inChina.
Comparedwiththesamplesizeofthepopulationinpre-viousresearches[24,27],ourpopulationwasmorecom-prehensivethanothersandwaslargerthan500,whichwassufficientforstatisticalpower[22].
ThepopulationpanelincludedcultivarsfromfivemainrepresentativeNieetal.
BMCGenomics(2016)17:352Page8of16cultivatedcottonregionsandwasthusenrichedwithabundantvariationsinyield,fiberqualityanddiseaseresistance.
Inthisstudy,theevaluationofsixfiberquality-relatedtraitsineightenvironmentsshowedwidevariations(0.
80~9.
31%),stableheritability(0.
84~0.
93)(Additionalfile7),andstablechangingtrendsofeachtraitindifferentenvironments(Fig.
4).
PhenotypictraitsanalysisbasedontheBLUPresultsruledoutenviron-mentaleffectsandimprovedtheaccuracyofthecom-plexquantitativetraits.
Boththecompositionofthepopulationandthetraitevaluationindicatedthatthispopulationpanelcouldbeconsideredasanidealre-sourceforassociationmappingofquantitativetraitsinG.
hirsutum.
MoleculargeneticdiversityInthisstudy,atotalof179SSRmarkersproduced426allelelociafterpolymorphismdetectionfrom494genome-wideSSRsinthepopulation.
Theaveragegen-eticdiversityandPICwere0.
377(rangingfrom0.
012to0.
893)and0.
336(rangingfrom0.
012to0.
887)respect-ively,whichwerehigherthanthatofQinetal.
[24].
Ourresearchresultsindicatedthattheselectedmarkershadsufficientpolymorphicinformationtorevealthegeneticrelationshipbetweentheseuplandcottonin-bredcultivars.
Theresultsalsoindicatedthattheaveragegeneticsimilaritycoefficientvariationwas0.
552(rangingfrom0.
337to0.
921),whichcouldbenefitcottonbreedersinselectingparentsinhybridizationbreedingtocreatenovelvariationsandtodevelopnewcultivars.
Compari-sonwithpreviouslypublishedstudies[5–9,28]showedthatthemoleculardiversityinourpopulationwashigher.
Infact,thepreviousresearchersemphasizedthenarrowgeneticbasisofuplandcottonandselectedonlyuplandcottongermplasmsrepresentingpartofthecot-tonplantedareasorbreedingperiods.
Forexample,thestudiesonuplandcottongermplasmsbyAietal.
[9]andNieetal.
[28]onlystudiedtheNorthwestInlandcottonvarieties(inthenorthandsouthofXinjiang),whichweremainlyderivedfromtheformerSovietUnion.
PopulationstructureTheevaluationofthepopulationgeneticstructureisaprerequisiteofgenome-wideassociationstudies[22],becausefalseassociationsusuallycausedbypopulationstructure[29].
Areasonablyaccuratepopulationstruc-turecanleadtomoregeneticsimilaritywithineachgroupandhighergeneticdifferencesamonggroupsandlargelyreducethedefectivesinassociationanalysis[30].
Thus,theaccuracyoftheassociationanalysisdependedonwhetherthepopulationstructurewasappropriate[31,32].
Previously,Abdurakhmonovetal.
usedQ-matrixtoestimateaccessionclusters[23].
Qinetal.
[24]andCaietal.
[33]determinedthepopulationstructureusingSTRUCTUREsoftwaretoshowKvalues.
Wangetal.
[34]determinedthegeneticstructureof55accessionsofG.
barbadensebyseparatingPCAplotsbasedonSSRandSRAPmarkers.
Inthisstudy,thethreemethodsofPCA,unrootedandroundedtree,withK=7corre-spondingtotheuppermoststructurallevelinthecon-tactzonemodel,werecombinedtopredictthepopulationstructurebasedon426allelelocigeneratedby179SSRmarkersdistributingonthe26chromo-somes.
Thepopulationwasclassifiedinto7subpopula-tions,whichwasreasonabletoeliminatethespuriousassociationeffectsintheassociationanalysis.
LinkagedisequilibriumLD,definedasnonrandomcombinationalallelesatdif-ferentgeneloci[29],isthegeneticunderpinningsofas-sociationanalysis.
Inourstudy,itwasshownthatdifferentnumbersoflineararraysoflinkagedisequilib-riumlocidistributedoneachchromosome,especiallythepair-locinumberofLDonChr01,10,12,15,19,21,and26(Fig.
3),onwhichthereweremorepolymorphiclociassociatedwithfiberqualitytraits.
Forexample,amongthe13polymorphiclocionChr01,8markerlociwererelatedtofiberqualitytraits.
Therefore,linkagedisequilibriumwasthebasisoftheassociationanalysis,inwhicheachpair-lociofLDrepresentedagroupallelicvariation.
ThestrongerLDdegreebetweenthelociinthedifferentallelicvariationswithfunctionaldifferences,themoreQTLassociatedwiththeirphenotypeweredetected.
RecombinationwasthegreatestfactorinfluencingLD.
Generally,LDlevelishighinself-pollinatedcrops,andlowincross-pollinationcropsbecauseitwillhaveahigherrecombinationrateandbreakthelinkagedisequi-libriumbetweengeneloci[35,36].
Cottonisacross-pollinationcropwithahigherrecombinationrate;more-over,asignificantamountofhumanbehavior,includingcross,backcross,openpollination,germplasmintroduc-tionandexchangeduringbreedingindifferentcottonplantingregionsgreatlyincreasetherecombinationrate,whichleadstothelowLDlevelinthecottongenome.
Abdurakhmonovetal.
[25]foundthattheattenuationdistanceofLDwas25cM,10cMand30cMatr2≥0.
1,andthedistancereached5–6cM,1–2cMand6–8cMatr2≥0.
2;Zhangetal.
[37]reportedthattheattenuationdistanceofLDwas3.
4cMandlessthan1cMatr2≥0.
1andr2≥0.
2respectively.
Fangetal.
[38]observedthattheattenuationdistanceofLDoftheaveragechromo-somewasapproximately6.
75cM.
Qinetal.
[24]discoveredthattheattenuationdistanceoflinkagedis-equilibriumdecreaseddramaticallyto0–10cM.
Inthisstudy,theattenuationdistanceofLDdecreaseddramat-icallyto0–5cM,asseenfromtheLDattenuationfigureNieetal.
BMCGenomics(2016)17:352Page9of16(Additionalfile5),whichwouldbeusefultofinemap-pingcandidateQTL.
AssociationmappingQTLforfiberqualitytraitsThe216markerloci(P<0.
05)associatedwithsixfibertraitsinthisstudywerecomparedtootherreportedQTLincotton.
Thirteenmarkerlociidentifiedinourstudycoincidedwithpreviousresearch,ofwhich3markerlociweredetectedwiththesametraits(Table2).
NAU3419[39]onChr02andrelatedtoFSwaslocatedatdifferentpositions.
NAU5480[40],dramaticallysig-nificantforFUHMLandwithmuchhigherexplanationofvariation,wasdetectedonChr11.
BNL3436[33]onChr25,significantforFS,wasdetectedbytwopreviousresearchers.
ThecommonQTLdetectedbydifferentre-searchersindicatedthatthemarkerlociassociatedwithfiberqualitytraitsinourstudywerereliable.
ComparingourQTLreferencesequencestoG.
arboreum[41],G.
raimondii[42]andG.
hirsutum[43]showedthat7QTLwererelatedtothegenefunctionoffiberdevelop-ment(Table3).
Recently,Shanetal.
verifiedthatahomeodomain-likesuperfamilyprotein,GhHOX3,couldcontrolcottonfiberelongation[44],andwasassociatedwiththemarkerlocusforfiberlengthanduniformity[45].
Coincidentally,somegenesannotatedasthehomeodomain-likesuperfamilyinthemarkerlocusHAU0211werefoundforfiberuniformityinourstudy.
AgeneannotatedasUDP-glycosyltranserasewasfoundtoberelatedtoBNL3436.
UDP-glycosyltransferasewasre-portedtobeinvolvedintheregulationofcellwallpectinbiosynthesis[46].
ThesecandidategeneswithintheQTLcouldaffectmaturefiberqualitytraitsbyregulatingfibercelldevelopmentandcellwallbiogenesis.
NAU6627(r2=1.
89,P=0.
0110)andMON-CGR5602(r2=1.
89,P=0.
0366),whichwere6.
7cMingeneticdistance,weresignificantlyassociatedwithFSonChr21.
NAU6627,relatedtoageneannotatedasmyosin2,acytoskeletonrelatedgene,waswithinthemarkerlocusBNL3436forTable2ComparisonofQTLassociatedwithfiberqualitytraitstothereportedstudiesLocusOurresearchOtherresearchPosition(cM)P-FDRr2TraitTraitCIR38138.
8684.
81E-020.
89FSLI[61];FE,FL,FS,MIC,FE,MAT,RD,FB[33];FL,MIC,MAT,FR,FB[62];PER,WF,SL[63]NAU3419156.
0344.
20E-021.
00FSFL,FS,FM[38];PB[64]NAU2836156.
2283.
40E-021.
08FUHMLFM[65]8.
69E-043.
17FSNAU5129140.
3223.
92E-021.
06FEBS[66]BNL379073.
21.
59E-021.
53FSSL[67]NAU5480173.
9787.
33E-043.
11FUHMLFS[68];FL,FS[40]4.
85E-021.
12FUNAU267140.
9953.
83E-021.
01FUHMLSY,LY,BPP[69]2.
42E-021.
32FS3.
29E-021.
48SF3.
87E-021.
26FUNAU646828.
4874.
93E-020.
88FUHMLFBN[61]0.
039801.
11MVHAU011902.
41E-021.
30FSFM[65]3.
46E-021.
39SFBNL334743.
2082.
34E-021.
34FSHP[67];MIC,FU,SCI[70]NAU3092236.
3224.
71E-020.
59FUHMLLI[64]0.
037690.
77MVGh27797.
1563.
87E-021.
4SFBNL343603.
45E-020.
78FSFL,FM[71]NAU308423.
9012.
68E-021.
39FUHMLFB[66];BW,NB[64]1.
59E-021.
51FS7.
34E-031.
93FU5.
12E-065.
08SFNieetal.
BMCGenomics(2016)17:352Page10of16fiberstrength.
Therewasareportthatanothercellskel-etoninteractionprotein,GhWLIM1a,couldpromoterfibercellelongationandregulatesfibersecondarycellwallbiogenesis[47].
OverexpressionoftheGhWIM1agenealsocouldincreasematurefiberstrength,whichindicatedthatmyosin2wasanimportantcandidategeneidentifiedfromQTLassociatedwithfiberstrength.
STV106wasre-latedtoanNACdomaincontainingprotein[48],whichcouldregulatesecondarywallbiogenesisandmaypartiallyexplainwhySTV106wasassociatedwithfivemainfiberqualitytraits.
CottonfibercellinitiationwassimilartotrichomeinArabidopsis,whichisregulatedbyWD40,bHLHandMYBtranscriptfactors[49].
WefoundthatsomegenesthatwereannotatedasWD40orbHLHinHAU1355,MONCGR5167andNAV2564.
Therelation-shipsbetweenfiberdevelopment-relatedgenesandma-turefiberqualitytraitsrequirefurtherresearch.
ApplicationsinbreedingTheinformationofthisstudyprovidedthecharacteris-ticsofphenotypicvariationoffiberqualitytraits,thegeneticdiversity,populationstructureandeliteallelesandencouragedustotakeafurtherstudytoproposeadetailedschemeforapplyingtheaboveresultsincottonbreeding.
Firstly,specificandeliteallelesconsideredasselectiontagsofgeneticfragmentsofintrogressionlineswereusedtocharacterizedifferentforegroundselectionpar-entsandwerethencrossedwithininfiltrationlinestoas-sembleeliteallelesfromdifferentmarkerlociintoonerecipientparentandbreedexcellenthybridprogenybyallelicbandsassistedselection.
Forexample,inourstudy,NAU5480(r2=3.
11),MON_DPL0544b(r2=1.
84)andHAU1166(r2=1.
82)associatedwithFUHML,NAU2836(r2=3.
17)andNAU2317(r2=1.
64)associatedwithFS,andDPL0457(r2=1.
46)andNAU6468(r2=1.
11)associatedwithMV.
Theseeliteallelescouldbecombinedbymarker-assistedselectiontodevelopacces-sionswithsuperfiberquality.
Secondly,elitegermplasmscouldbeselectedaspar-entsinthebreedingprogrambasedontheassociationresults;thetypicalcarriermaterialsaggregatedtheallelicvariationwiththemostpositiveefficiency,moredistantgeneticrelationshipandcomplementaryeliteandstablegenetictraits.
Forexample,accordingtothephenotypiceffectsofeachgermplasmbasedontheBLUPofsixfiberqualitytraitsineightenvironments(Additionalfile6),ZY495,ZY488andZY64hadhigherphenotypiceffectsforFUHML,FSandMV,respectively;meanwhile,ZY43hadhigherphenotypiceffectsforFUHMLandrelativelylowerphenotypiceffectsforFS.
ComparingthegeneticsimilaritycoefficientsofZY495,ZY488andZY64toZY43,thelowestgeneticsimilaritycoefficientbetweenZY64andZY43indicatedthattheycouldbeselectedtoimprovefiberquality.
ConclusionsAnassociationmappingpopulationof503nation-wideuplandcottonsinChinawasgenotypedby494genome-wideSSRsandphenotypedineightenviron-ments,whichrevealedabundantmoleculardiversityandphenotypicvariations.
ThepopulationwasdividedTable3AllelicvariationlociassociatedwithfiberqualityandannotatedgenesAllelicvariationlociChromosomeTraitPosition(cM)P-FDRr2GeneannotationHAU021112FU10.
1923.
77E-021.
24Homeodomain-likesuperfamilyproteinSF4.
48E-021.
14HAU135518FUHML118.
0423.
26E-020.
76WD40/YVTNrepeat-like-containingdomainFS1.
72E-021.
08SF3.
40E-020.
96MON_CGR516711FUHML61.
2063.
88E-021.
04basichelix-loop-helix(bHLH),DNA-bindingsuperfamilyproteinBNL343625FS03.
45E-020.
78UDP-glycosyltransferase73B4NAU25647FUHML18.
2233.
05E-021.
17basichelix-loop-helix(bHLH),DNA-bindingsuperfamilyproteinSF2.
71E-021.
63FS3.
53E-032.
33NAU662721FS90.
4341.
10E-021.
89myosin2STV1066FUHML01.
88E-021.
59NACdomaincontainingprotein28MV0.
037751.
19FU4.
59E-021.
17SF2.
77E-021.
75FE4.
38E-021.
06Nieetal.
BMCGenomics(2016)17:352Page11of16intosevensubpopulationsbycomprehensiveanalysis;andtheattenuationdistanceofLDinthispopulationwas0–5cM.
MLMbasedassociationmappingforfiberqualitydetectedknownandnovelelitealleles,andtypicalmaterialswereidentified.
Theresultsinthisstudywillprovideaplatformforfuturegeneticsandbreedingincotton.
MethodsPlantmaterialsAsetof503uplandcottoninbredcultivars(Additionalfile3)constructedthemappingpopulationpanel,whichrepresentsdiversegeneticresourcesrelatedtofiberqual-itytraits.
TheseinbredcultivarsweremainlycollectedfromChinaandtwoforeigncountriesandweredividedintosevencottongrowingregionsaccordingtodifferentecologicalcharacteristics,including79fromtheNNIR,28fromtheNSEMR,225fromtheYRR,141fromtheYtRR,4fromtheSC,20fromtheUnitedStates,and6theSovietUnion.
TheseinbredcultivarsarefrontierspecimensforcultivardevelopmentinChina.
Thecultivarsinthepopulationpanelcoveredfourvarietyimprovementstages:thepreliminaryphaseoftheG.
arboreumcottonanduplandcottonimprovement(1920s–1950s);,thebreedingstageofincreasingyield(late1950s–1970s);thebreedingstageofcomprehen-siveimprovementofyield,fiberqualityanddiseaseresistance(1980s–1990s);andthebreedingstageofgoodfiber,geneticallymodifiedandmechanizedhar-vestingcottonvarieties(1990s–2010s).
The503uplandcottoncultivarswerecollectedfromtheInstituteofCottonResearchofChineseAcademyofAgriculturalSciences(Anyang,China),XinjiangAcademyofAgricultureandReclamationScience(Shihezi,China),ShiheziAcademyofAgriculturalSciences(Shihezi,China),HenanAcademyofAgriculturalSciences(Zhengzhou,China),JiangsuAcademyofAgriculturalSciences(Nanjing,China),CottonResearchInstituteofShanxiAacademyofAgriculturalSciences(Yuncheng,China),ShandongCottonResearchCenter(Jinan,China),andHebeiAgriculturalUniversity(Baoding,China).
Allthecultivarswereauthorizedforonlyscientificresearchpurpose,andweredepositedintheoriginalinstitutesandHuazhongAgriculturalUniversity.
FieldexperimentsandphenotypedatacollectionFieldexperimentswereperformedinShihezi(SHZ),NorthXinjiang(E85.
94°,N44.
27°,theNIR),Kuerle(KEL),SouthXinjiang(E86.
06°,N41.
68°,theNIR),Yuanyang(YY),Henan(E113.
97°,N35.
05°,theYRR),andHuanggang(HG),Hubei(E114.
87°,N30.
44°,theYtRR)in2012and2013.
Theselocationsareincludedinthemaincultivatedcottonareascoveringthecot-tonecologicalareas.
Theuplandcottoncultivarswecollectedarenotendangeredorprotectedspecies.
Thefieldworksandsamplingwereauthorizedbylocalgovernments,respectively.
Standardfieldplotsandagronomictechnologieswereusedtogrowthepopulationsineightenviron-ments(intwoyearsandfourlocations);andfieldtri-alsfollowedacompletelyrandomizedblockdesignwithtworeplicatesineachenvironment.
Takingintoconsiderationofthedifferentcultivationmanagementpatternsinfourlocations,thedesignsofthefieldex-perimentswereconductedasfollows:usingaspecificwide-narrowdistanceplantingpattern,therowspa-cingwas(40+50+46)cm,with9.
5cmbetweenindi-vidualsand50individualsper5minonerow,plantedbysowinginaholeanddrippingirrigationinXinjiang(bothinShiheziandKuerle);therowspa-cingwas90cm,with30cmbetweenindividualsand12individualsper5minonerow,plantedbysowinginholeandfloodinginYuanyang;therowspacingwas100cm,with40cmbetweenindividualsand10individualsper5minonerow,seedlingtransplant-ationandwithoutirrigatinginHuanggang.
Tomeasurethefiberqualitytraits,20bollswerecollectedfromthemiddlefruitbranchesofplantsineachline.
Aftergrinning,10–15gfibersofeachsam-pleweresenttotheInstituteofCottonResearch,ShiheziAcademyofAgriculturalSciencestotestthefiberquality.
Thefiberquality-relatedtraitsweretestedat20°C,and65%relativehumiditywithanHVI1000AutomaticFiberDeterminationSystem,in-cludingthefiberupperhalfmeanlength(FUHML),fiberstrength(FS),micronairevalue(MV),fiberuniformity(FU),shortfiber(SF),andfiberelong-ation(FE).
PhenotypedataanalysisThefiberqualitytraitsstatisticswereobtainedfromone-year-one-locationandeightenvironmentsbybestlinearunbiasedprediction(BLUP)[50].
Broad-senseher-itabilitywascalculatedforeachtraitaccordingtothemethoddescribedbyLiuetal.
[51],whichreferstothepercentageofthetotalvarianceofphenotypeofgeneticvariation,orphenotypevarianceofgeneticvariancepercentage.
Analysisofphenotypicchangingtrendsandrelevanceoffiberqualitytraitswereshowninboxplotform,andthecorrelationpicturesbetweentheenvironmentandenvironmentbythedifferentcodeweredrawnusingthe"R"program.
SPSS17.
0software(http://www-01.
ibm.
com/software/analytics/spss/)wasusedtocalculatebasicstatistics,in-cludingextremevalues,meanvalue,standarddeviation,variablecoefficient,andthestatisticsforthecorrelationbetweentraits.
Nieetal.
BMCGenomics(2016)17:352Page12of16SSRmarkersgenotypingandgeneticdiversityAtotalof494SSRmarkerswereselectedatanaverageof10cMfromtheinterspecificgeneticmapconstructedwith5152markersincotton[12]andwereusedtoscreenthegeneticpolymorphismofthepopulation.
TotalDNAwasextractedfromyoungleavesusingthemodifiedCTABmethod[52].
Poly-merasechainreaction(PCR),electrophoresis(6%denaturingpolyacrylamidegelelectrophoresisforSSRand8%non-denaturingpolyacrylamidegelelectrophoresisforSSCP)andsilverstainingwereperformedaccordingtothemethodsdescribedbyLietal.
[52].
SSRandSSCPfragmentswerecodedas"1"forpresent,"0"forabsent,and""formissingdata.
Thenumberofallelesloci,numberofgenotypes,genediversity,andpolymorphisminformationcontent(PIC)ofpolymorphicmarkerswerecalculatedbyPowermar-kersoftware3.
25[53](http://tree.
bio.
ed.
ac.
uk/software/figtree/).
Geneticdifferentiation(PhiPT)wasestimatedusingtheAMOVAandtheFrequencyfunctionofGENALEX6.
2software[54](http://biology.
anu.
edu.
au/GenAlEx/).
ThehierarchicaldiagramandmatricesofgeneticsimilaritywereobtainedfromtheNTSYS-pcversion2.
10estatisticalpackagebasedonJaccard'salgo-rithms[55].
PopulationstructureUnrootedtreeandcircularclusteringfigureweregener-atedonthebasisofthegenefrequency,geneticdistance,andanalysisofneighbor-joining(NJ)treesfromPower-markersoftware3.
25[53].
Atwo-dimensionaldiagramofprincipalcoordinatesanalysis(PCA)wasproducedbasedonageneticdistance(GD)matrixbyGENA-LEX6.
2software[54].
Thepopulationstructurewasanalyzedandevalu-atedforfurtherassociationmappingbySTRUC-TUREversion2.
1[31,32]usinganadmixturemodel;MCMCwassetto50000,burn-inwassettoarunningtimeof100000andtheKvaluewassetto2–10.
EachKvaluewasrepeatedfivetimes.
Tose-lectasuitableK,drawingLnP(D)andKvalues-changingtrenddiagramswereusedtodeterminewhichKvaluereachedthelargestplateauoftheLnP(D)values.
Iftherewasnosuchplateau,thenthedrawingdelta(K)andKvalues-changingtrenddia-gramswereusedtojudgewhichinflectionpointoftheKvaluewasthemostsuitableforthenumberofthepopulation[56].
Clusteranalysis,principalcom-ponentanalysis(PCA)andstructureanalysiswereusedtodetermineareasonablepopulationstructureandcorrectfalsepositivescausedbythepopulationstructureforassociationmappinginthisstudy.
LinkagedisequilibriumandLDdecayThedistributiondiagramsofthegenome-widelinkagedisequilibrium(LD)attenuationweredrawninSPSSsta-tistics17.
0software(http://www-01.
ibm.
com/software/analytics/spss/)basedonr2andD'fromtheresultsoftheLDrunninginTASSEL[57]andthegeneticdis-tance.
TheLDdecayscatterplotshowstherelationshipbetweenr2andD'onthey-axisandthegeneticdistancebetweenmarkerpairsonthex-axistounderstandthepatternofLDinthevariousannotationgroupsorentiregenome[58].
AssociationanalysisTheexistenceofthepopulationstructureandrelativekinshipinthepopulationalwaysresultinahighlevelofspuriouspositivesinassociationmapping.
Inthisstudy,theTASSELV2.
1[57]softwarepackagewasusedtoanalyzethemolecularmarkerdataandthephenotypicdataofthefiberqualityineightdifferentenvironmentsindividuallyusingthethreemodelsofGLM(P+G+QandP+G+PCA)andMLM(P+G+Q+K).
TheQmatrixwasderivedfromClumppsoftware[59]tomergefiverepetitiveresultsuptotheKvalueforsevenusingtheStructuresoftwarepackage.
KingshipwasgeneratedfromtheTASSELsoftwarepackageresults.
ThePvaluesofmarkersassociatedwithQTLwereregulatedbythemethodofmultipletestingcorrectionbycontrollingthefalsediscoveryrate[60].
Elitealleleexplorationandgene-relatedfunctionalannotationAccordingtotheresultsofthegenome-wideanalysisstudy,theprimersequencesoftheQTLthatweremappedonasinglechromosome,repeatedinmul-tipleenvironmentsandreportedthesameaspreviousstudieswereconsideredtocomparewiththeG.
arboreum[41],G.
raimondii[42]andG.
hirsutum[43]genomesequences,andtheninternalgenespacerandgeneannotationincludingthefunctionandpath-waywerediscovered.
Ethics(andconsenttoparticipate)Notapplicable.
ConsenttopublishNotapplicable.
AvailabilityofdataandmaterialsAllrelevantdataareavailablewithinthemanuscriptanditsadditionalfiles.
Nieetal.
BMCGenomics(2016)17:352Page13of16AdditionalfilesAdditionalfile1:TableS1.
TheSSRmarkersinformationof503cultivars.
(XLSX22kb)Additionalfile2:FigureS1.
Clusteringanalysistreeof503G.
hirsutumcultivarsbasedonJaccard'ssimilaritycoefficientsbySSRs.
(PDF27kb)Additionalfile3:TableS2.
Thepedigreesorigininformationof503cultivars.
(XLS124kb)Additionalfile4:FigureS2.
Neighbor-joiningtreesof503G.
hirsutumcultivars.
Dendrogramof503G.
hirsutumcultivarsbyNJanalysisintheformofunrootedtree.
(a):BasedonNei'sgeneticdistance,thepopulationformed7distinctgroupsfromItoVII,including106,19,147,67,9,103,and52cultivars,respectively;(b):Dendrogramof503G.
hirsutumcultivarsbyNJanalysisintheformofroundtree.
Colorsinthedendrogramcorrespondtothepopulationfromsevendifferentcottonregions.
(TIF5688kb)Additionalfile5:FigureS3.
LDdecaysasafunctionofgeneticdistanceintheassociationpanelconsistingof503uplandcottoncultivars.
(a):LDdecayswithindistanceintheformofr2basedonlinkagedisequilibriumlocionthesamechromosome;(b):LDdecayswithdistanceintheformofr2basedondifferentintervalsoflinkagedisequilibriumlocionthesamechromosome;(c):LDdecayswithdistanceintheformofD'basedonlinkagedisequilibriumlocionthesamechromosome.
(TIF2615kb)Additionalfile6:TableS3.
TheBLUPofsixfiberqualitytraitsof503cultivars.
(XLS243kb)Additionalfile7:TableS4.
Phenotypestatisticsofsixfiberqualitytraitsinthetestedcottoncultivars.
(XLSX12kb)Additionalfile8:TableS5.
Thecorrelationanalysisofsixfiberqualitytraits.
(XLSX10kb)Additionalfile9:FigureS4.
Quantile–quantileplotsofestimated–log10(P-value)fromassociationanalysisusingthreemodels.
(a)FUHML;(b)FS;(c)MV;(d)FU;(e)SF;(f)FE.
Theblacklineistheexpectedlineunderthenulldistribution.
Undertheassumptionthattherearefewtruemarkerassociations,theobservedPvaluesareexpectedtonearlyfollowtheexpectedPvalues.
(TIF8527kb)Additionalfile10:TableS6.
Themarkerlocirelatedtosixfiberqualitytraitsofphenotypicvariation.
(XLSX18kb)Additionalfile11:TableS7.
TheresultsofcomparisonoftheallelicvariationlocirelatedfiberqualityandtheGossypiumarboreumgenome.
(XLSX14kb)Additionalfile12:TableS8.
Theresultsofthecomparisonoftheallelicvariationloci-relatedfiberqualityandtheGossypiumraimondiigenome.
(XLSX14kb)Additionalfile13:TableS9.
Theresultsofthecomparisonoftheallelicvariationloci-relatedfiberqualityandtheGossypiumhirsutumgenome.
(XLSX14kb)Additionalfile14:TableS10.
TheSSRmarkerlociassociatedwithsixfiberqualitytraitsofphenotypiceffectsandtypicalmaterials.
(XLSX15kb)AbbreviationsBLUP:bestlinearunbiasedprediction;FE:fiberelongation;FS:fiberstrength;FU:fiberuniformity;FUHML:fiberupperhalfmeanlength;LD:linkagedisequilibrium;MLM:mixedlinearmodel;MV:micronairevalue;NIR:NorthwesternInlandRegion;NSEMR:NorthernSpecificEarlyMaturationRegion;QTL:quantitativetraitloci;SCR:SouthChinaRegion;SF:shortfiber;SSR:simplesequencerepeat;YRR:YellowRiverRegion;YtRR:YangtzeRiverRegion.
CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.
Authors'contributionXHNconductedtheexperimentswithCH;WXZandCSconductedassociationanalysisandco-wrotethemanuscripttext.
ZXLdesignedandsupervisedtheresearchwithXLZandrevisedthemanuscript.
XHN,CYY,WL,CH,WXZ,CS,BBZ,HTW,ZHYandBSDinvestigatedphenotypictraitsinKuerle,Shihezi,YuyangandHuanggang(in2012and2013,respectively),andconducteddataanalysis.
MJWanalyzedtheQTLrelatedgenesinthecottongenome.
Allauthorsreviewedthemanuscript.
Alltheauthorshadreadandapprovedthefinalversionofthemanuscript.
Authors'informationXHN,CH,WXZ,CS,BBZ,HTW,ZHY,BSD,MJW,XLZ,andZXL:NationalKeyLaboratoryofCropGeneticImprovement,CollegeofPlantScience&Technology,HuazhongAgriculturalUniversity,Wuhan430070,Hubei,China;XHN,andCYY:CottonResearchInstitute,ShiheziAcademyofAgricultureScience,Shihezi832011,Xinjiang,China.
WL:EconomicCropResearchInstitute,HenanAcademyofAgriculturalSciences,Zhengzhou450002,Henan,China.
AcknowledgementsWewouldliketothanktheInstituteofCottonResearchofCAAS(Anyang,Henan),YuYu(CottonInstitute,XinjiangAcademyofAgricultureandReclamationScience,Shihezi,Xinjiang),ChunyuanYou(CottonInstitute,ShiheziAcademyofAgriculturalSciences,Xinjiang),WuLi(HenanAcademyofAgriculturalSciences,Zhengzhou,Henan),XinlianShen(EconomicCropsResearchInstitute,JiangsuAcademyofAgriculturalSciences,Nanjing,Jiangsu),YanliJiang(CottonResearchInstituteofShanxiAacademyofAgriculturalSciences,Yuncheng,Shanxi),ZhiweiWang(ShandongCottonResearchCenter,Jinan,Shandong),andZhikunLi(HebeiAgriculturalUniversity,Baoding,Hebei)whoprovideduscottoncultivarseeds.
WealsowouldliketothankYuYu,JilianLi,XinzhuNing,JunjiSu(CottonInstitute,XinjiangAcademyofAgricultureandReclamationScience,Shihezi,Xinjiang),JianghongQin(CottonInstitute,ShiheziAcademyofAgriculturalSciences,Xinjiang),andWuLiandDeyiXie(HenanAcademyofAgriculturalSciences,Zhengzhou,Henan)whoprovidedusexperimentalfieldsandinvestigatedthephenotypictraitsinKuerle,ShiheziandYuyang.
FundingThisworkwasfinanciallysupportedbytheFundamentalResearchFundsfortheCentralUniversities(GrantNo.
2014PY015).
Authordetails1NationalKeyLaboratoryofCropGeneticImprovement,CollegeofPlantSciences&Technology,HuazhongAgriculturalUniversity,Wuhan430070,Hubei,China.
2CottonResearchInstitute,ShiheziAcademyofAgricultureScience,Shihezi832011,Xinjiang,China.
3EconomicCropResearchInstitute,HenanAcademyofAgriculturalSciences,Zhengzhou450002,Henan,China.
Received:9November2015Accepted:23April2016References1.
ChenZJ,SchefflerBE,DennisE,TriplettBA,ZhangT,GuoW,etal.
Towardsequencingcotton(Gossypium)genomes.
PlantPhysiol.
2007;145:1303–10.
2.
JiDP.
CottonEncyclopedicKnowledge.
Beijing:ChinaAgriculturalPress;2001.
3.
ZhouSH.
GenogramofcottonvarietiesinChina.
Sichuan:SichuanScienceandTechnologyPress;2000.
4.
HuangZK.
CottonvarietiesandtheirgenealogyinChina.
Beijing:ChineseAgriculturalPress;2007.
5.
XuQH,ZhangXL,NieYC.
StudyoncomparisonofgeneticdiversitybetweentheYangtzeRiverandYellowRivervalley.
ActaGenetSin.
2001;28:683–90.
6.
WenXJ,MaCY,WangSF,ZhangGY,LiXH.
MoleculardivergenceestimationofChineseuplandcottonswithfusariumandverticilliumwiltResistance.
ActaAgronSin.
2005;38:936–43.
7.
WuDP,FangXX,CuiRG,ChenJH,ZhuSJ.
GeneticrelationshipanddiversityoftheuplandcottongermplasmsfromdifferentcottonproducingcountriesusingSSRmarkers.
CottonSci.
2011;23:291–9.
8.
ZhaoZS,YuSX,FanSL,ZhangWG,PangCJ,MeiYJ.
Analysisofgeneticdiversityofearlymaturinguplandcottonvarietiesinnorthern.
CottonSci.
2012;24:473–80.
Nieetal.
BMCGenomics(2016)17:352Page14of169.
AiXT,LiangYJ,ShaH,WangJZ,ZhengJY,TuEXJ,etal.
GeneticdiversityanalysisonlocaluplandcottoncultivarsinXinjiangbasedonSSRmarkers.
ActaAgronSin.
2014;40:369–79.
10.
KongFL.
Quantitativegeneticsinplants.
Beijing:ChinaAgriculturalUniversityPress;2005.
11.
YuY,YuanDJ,LiangSG,LiXM,WangXQ,LinZX,etal.
Genomestructureofcottonrevealedbyagenome-wideSSRgeneticmapconstructedfromaBC1populationbetweenGossypiumhirsutumandG.
barbadense.
BMCGenomics.
2011;12:15.
12.
LiXM.
Constructionofintrogressionlines,developmentofnewmarkers,andconstructionofhigh-densitygeneticlinkagemapincotton.
HuazhongAgriculturalUniversitylibrary.
2013.
13.
LiuRZ,AiNJ,ZhuXX,LiuFJ,GuoWZ,ZhangTZ.
Geneticanalysisofplantheightusingtwoimmortalizedpopulationsof"CRI12*J8891"inGossypiumhirsutumL.
Euphytica.
2014;196:51–61.
14.
NingZY,ChenH,MeiHX,ZhangTZ.
MoleculartaggingofQTLsforfiberqualityandyieldintheuplandcottoncultivarAcala-Prema.
Euphytica.
2014;195:143–56.
15.
FangDD,JenkinsJN,DengDD,McCartyJC,LiP,WuJX.
Quantitativetraitlocianalysisoffiberqualitytraitsusingarandom-matedrecombinantinbredpopulationinUplandcotton(GossypiumhirsutumL.
).
BMCGenomics.
2014;15:397.
16.
AiNJ,ZhuXX,GuanRZ,ZhaoJJ,ZhangTZ.
Geneticanalysisofmajorlocusgroupconstitutionsofgrowthstagesinuplandcotton.
SciAgricSin.
2010;43:4140–8.
17.
BolekY,El-ZikKM,PepperAE,BellAA,MagillCW,ThaxtonPM,etal.
Mappingofverticilliumwiltresistancegenesincotton.
PlantSci.
2005;168:1581–90.
18.
WangHM,LinZX,ZhangXL,ChenW,GuoXP,NieYC,etal.
Mappingandquantitativetraitlocianalysisofverticilliumwiltresistancegenesincotton.
JIntegrPlantBiol.
2008;50:174–82.
19.
JiangF,ZhaoJ,ZhouL,GuoWZ,ZhangTZ.
MolecularmappingofverticilliumwiltresistanceQTLclusteredonchromosomesD7andD9inuplandcotton.
SCIChinaSerC.
2009;52:872–84.
20.
FangH,ZhouHP,SanogoS,FlynnR,PercyRG,HughsSE,etal.
Quantitativetraitlocusmappingforverticilliumwiltresistanceinabackcrossinbredlinepopulationofcotton(Gossypiumhirsutum*Gossypiumbarbadense)basedonRGA-AFLPanalysis.
Euphytica.
2013;194:79–91.
21.
ZhaoYL,WangHM,ChenW,LiYH.
Geneticstructure,linkagedisequilibriumandassociationmappingofverticilliumwiltresistanceinelitecotton(GossypiumhirsutumL.
)germplasmpopulation.
PlosOne.
2014;9:e86308.
22.
YangXH,YanJB,ZhengYP,YuJM,LiJS.
Reviewsofassociationanalysisforquantitativetraitsinplants.
ActaAgronSin.
2007;33(4):523–30.
23.
AbdurakhmonovIY,SahaS,JenkinsJN,BurievZT,ShermatovSE,SchefflerBE,etal.
LinkagedisequilibriumbasedassociationmappingoffiberqualitytraitsinG.
hirsutumL.
varietygermplasm.
Genetica.
2009;136:401–17.
24.
QinHD,ChenM,YiXD,BieS,ZhangC,ZhangYC,etal.
IdentificationofassociatedSSRmarkersforyieldcomponentandfiberqualitytraitsbasedonframemapanduplandcottoncollections.
PlosOne.
2015;10:e0118073.
25.
AbdurakhmonovIY,KohelRJ,YuJZ,PepperAE,AbdullaevAA,KushanovFN,etal.
MoleculardiversityandassociationmappingoffiberqualitytraitsinexoticG.
hirsutumL.
germplasm.
Genomics.
2008;92:478–87.
26.
YuSX.
GeneticsandBreedingofCottoninChina.
Shandong:ShandongScienceandTechnologyPress;2003.
27.
NieXH,YouCY,BaoJ,LiXF,HuiH,LiuHL,etal.
ExplorationofeliteallelesofagronomicandfiberqualitytraitsinXinluzaocottonvarietiesbyassociationanalysis.
SciAgricSin.
2015;48(15):2891–910.
28.
NieXH,YouCY,LiXF,QinJH,HuangC,GuoHL,etal.
ConstructionofDNAfingerprintingandanalysisofgeneticdiversityforXinluzaocottonvarieties.
ActaAgronSin.
2014;40:2113–26.
29.
Flint-GarciaSA,ThornsberryJM,BucklerES.
Structureoflinkagedisequilibriuminplants.
AnnuRevofPlantBiol.
2003;54:357–74.
30.
WeiSP,LiuXF,YangSX,LuHY,NiuY,ZhangYM.
ComparisonofvariousclusteringmethodsforpopulationstructureinChinesecultivatedsoybean(Glycinemax(L.
)Merr.
).
JNanjingAgriUniv.
2011;34:13–7.
31.
PritchardJK,StephensM,DonnellyP.
Inferenceofpopulationstructureusingmultilocusgenotypedata.
Genetics.
2000;155:945–59.
32.
PritchardJK,StephensM,RosenbergNA,DonnellyP.
Associationmappinginstructuredpopulations.
AmJHumGenet.
2000;67:170–81.
33.
CaiCP,YeWX,ZhangTZ,GuoWZ.
AssociationanalysisoffiberqualitytraitsandexplorationofeliteallelesinUplandcottoncultivars/accessions(GossypiumhirsutumL.
).
JIntegrPlantBiol.
2014;56:51–62.
34.
WangXQ,YuY,LiW,GuoHL,LinZX,ZhangXL.
AssociationanalysisofyieldandfiberqualitytraitsinGossypiumbarbadensewithSSRsandSRAPs.
GenetMolRes.
2013;12(3):3353–62.
35.
GautBS,LongAD.
Thelowdownonlinkagedisequilibrium.
PlantCell.
2003;15:1502–6.
36.
ShifmanS,KuypersJ,KokorisM,YakirB,DarvasiA.
Linkagedisequilibriumpatternsofthehumangenomeacrosspopulations.
HumMolGenet.
2003;12:771–6.
37.
ZhangYC,BieS,Xian-DaYI,ZhangC,Cheng-QiLI,QinHD.
PopulationStructureandLinkagedisequilibriumanalysisofgermplasmresourcesinuplandcotton.
CottonSci.
2011;23:500–6.
38.
FangDD,HinzeLL,PercyRG,LiP,DengD,ThyssenG.
Amicrosatellite-basedgenome-wideanalysisofgeneticdiversityandlinkagedisequilibriuminuplandcotton(GossypiumhirsutumL.
)cultivarsfrommajorcottongrowingcountries.
Euphytica.
2013;191:391–401.
39.
CaoZB,WangP,ZhuXF,ChenH,ZhangTZ.
SSRmarker-assistedimprovementoffiberqualitiesinGossypiumhirsutumusingG.
barbadenseintrogressionlines.
TheorApplGenet.
2014;127:587–94.
40.
YuJW,ZhangK,LiSY,YuSX,ZhaiHH,WuM,etal.
MappingquantitativetraitlociforlintyieldandfiberqualityacrossenvironmentsinaGossypiumhirsutum*Gossypiumbarbadensebackcrossinbredlinepopulation.
TheorApplGenet.
2013;126:275–87.
41.
LiFG,FanGY,WangKB,SunFM,YuanYL,SongGL,etal.
GenomesequenceofthecultivatedcottonGossypiumarboreum.
NatGenet.
2014;46:567–72.
42.
PatersonAH,WendelJF,GundlachH,GuoH,JenkinsJ,JinDC,etal.
RepeatedpolyploidizationofGossypiumgenomesandtheevolutionofspinnablecottonfibres.
Nature.
2012;492:423–7.
43.
ZhangTZ,HuY,JiangWK,FangL,GuanXY,ChenJD,etal.
Sequencingofallotetraploidcotton(GossypiumhirsutumL.
acc.
TM-1)providesaresourceforfiberimprovement.
NatBiotechnol.
2015;33:531–U252.
44.
ShanCM,ShangguanXX,ZhaoB,ZhangXF,ChaoLM,YangCQ,etal.
ControlofcottonfibreelongationbyahomeodomaintranscriptionfactorGhHOX3.
NatCommun.
2014;5:5519.
45.
LuanM,GuoX,ZhangY,YaoJ,ChenW.
QTLmappingforagronomicandfibertraitsusingtwointerspecificchromosomesubstitutionlinesofUplandcotton.
PlantBreed.
2009;128:671–9.
46.
QinLX,RaoY,LiL,HuangJF,XuWL,LiXB.
CottonGalT1encodingaputativeglycosyltransferaseisinvolvedinregulationofcellwallpectinbiosynthesisduringplantdevelopment.
PlosOne.
2013;8:e59115.
47.
HanLB,LiYB,WangHY,WuXM,LiCL,LuoM.
ThedualfunctionsofWLIM1aincellelongationandsecondarywallformationindevelopingcottonfibers.
PlantCell.
2013;25:4421–38.
48.
WangHZ,ZhaoQ,ChenF,WangMY,DixonRA.
NACdomainfunctionandtranscriptionalcontrolofasecondarycellwallmasterswitch.
PlantJ.
2011;68:104–1114.
49.
PeschM,HülskampM.
One,two,threemodelsfortrichomepatterninginArabidopsisCurrOpininPlantBiol.
2009;12:587–92.
50.
TechnowF,RiedelsheimerC,SchragTA,SchragTA,MelchingerAE.
Genomicpredictionofhybridperformanceinmaizewithmodelsincorporatingdominanceandpopulationspecificmarkereffects.
TheorApplGenet.
2012;125:1181–94.
51LiuLF,MaSX,HuangZY.
Thecropquantitativegenetics.
Beijing:Agriculturepress;1984.
52.
LiXM,YuanDJ,WangHT,ChenXM,WangB,LinZX,etal.
IncreasingcottongenomecoveragewithpolymorphicSSRsasrevealedbySSCP.
Genome.
2012;55(6):459–70.
53.
LiuKJ,MuseSV.
PowerMarker:integratedanalysisenvironmentforgeneticmarkerdata.
Bioinformatics.
2005;21:2128–9.
54.
PeakallR,SmousePE.
GENALEX6:geneticanalysisinExcel.
Populationgeneticsoftwareforteachingandresearch.
MolEcolNotes.
2006;6:288–95.
55.
AdamsDC,RohlfFJ.
EcologicalcharacterdisplacementinPlethodon:biomechanicaldifferencesfoundfromageometricmorphometricstudy.
PNatlAcadSciUSA.
2000;97:4106–11.
56.
MezmoukS,DubreuilP,BosioM,DécoussetL,CharcossetA,PraudS,etal.
Effectofpopulationstructurecorrectionsontheresultsofassociationmappingtestsincomplexmaizediversitypanels.
TheorApplGenet.
2011;122:1149–60.
Nieetal.
BMCGenomics(2016)17:352Page15of1657.
BradburyPJ,ZhangZ,KroonDE,CasstevensTM,RamdossY,BucklerES.
TASSEL:softwareforassociationmappingofcomplextraitsindiversesamples.
Bioinformatics.
2007;23:2633–5.
58.
XiaoYJ,CaiDF,YangW,YeW,YounasM,WuJS,etal.
Geneticstructureandlinkagedisequilibriumpatternofarapeseed(BrassicanapusL.
)associationmappingpanelrevealedbymicrosatellites.
TheorApplGenet.
2012;125:437–47.
59.
JakobssonM,RosenbergNA.
CLUMPP:aclustermatchingandpermutationprogramfordealingwithlabelswitchingandmultimodalityinanalysisofpopulationstructure.
Bioinformatics.
2007;23:1801–6.
60.
BenjaminiY,HochbergY.
Controllingthefalsediscoveryrate:apracticalandpowerfulapproachtomultipletesting.
JRStatSocSerCApplStat.
1995;57:289–300.
61.
LiuRZ,WangBH,GuoWZ,QinYS,WangLG,ZhangYM,etal.
QuantitativetraitlocimappingforyieldanditscomponentsbyusingtwoimmortalizedpopulationsofaheterotichybridinGossypiumhirsutumL.
MolBreeding.
2012;29:297–311.
62.
WangBH,GuoWZ,ZhuXF,WuYT,HuangNT,ZhangTZ.
QTLmappingoffiberqualityinanelitehybridderived-RILpopulationofuplandcotton.
Euphytica.
2006;152:367–78.
63.
WuJX,GutierrezOA,JenkinsJN,McCartyJC,ZhuJ.
QuantitativeanalysisandQTLmappingforagronomicandfibertraitsinanRIpopulationofuplandcotton.
Euphytica.
2009;165:231–45.
64.
ZhangTZ,QianN,ZhuXF,ChenH,WangS,MeiHX,etal.
VariationsandtransmissionofQTLallelesforyieldandfiberqualitiesinUplandcottoncultivarsdevelopedinChina.
PlosOne.
2013;8:e57220.
65.
SunFD,ZhangJH,WangSF,GongWK,ShiYZ,LiuAY,etal.
QTLmappingforfiberqualitytraitsacrossmultiplegenerationsandenvironmentsinuplandcotton.
MolBreeding.
2012;30:569–82.
66.
GuoX,GuoYP,MaJ,WangF,SunMZ,GuiLJ,etal.
MappingheteroticlociforyieldandagronomictraitsusingchromosomesegmentintrogressionlinesinCotton.
JIntegrPlantBiol.
2013;55:759–74.
67.
AnCF,JenkinsJN,WuJX,GuoYF,McCartyJC.
UseoffiberandfuzzmutantstodetectQTLforyieldcomponents,seed,andfibertraitsofuplandcotton.
Euphytica.
2010;172:21–34.
68.
ZhangXJ,YuanYC,WeiZ,GuoX,GuoYP,ZhangSQ,etal.
MolecularmappingandvalidationofamajorQTLconferringresistancetoadefoliatingisolateofverticilliumwiltincotton(GossypiumhirsutumL.
).
PlosOne.
2014;9:e96226.
69.
ZhangW,LiuF,LiSH,WangW,WangC,ZhangX,etal.
QTLAnalysisonyieldanditscomponentsinrecombinantinbredlinesofuplandcotton.
ActaAgronSin.
2011;3:433–42.
70.
ShenXL,ZhangTZ,GuoWZ,ZhuXF,ZhangXY.
MappingfiberandyieldQTLswithmain,epistatic,andQTL*environmentinteractioneffectsinrecombinantinbredlinesofuplandcotton.
CropSci.
2006;46:61–6.
71.
ShenXL,GuoWZ,ZhuXF,YuanYL,YuJZ,KohelRJ,etal.
MolecularmappingofQTLsforfiberqualitiesinthreediverselinesinUplandcottonusingSSRmarkers.
MolBreeding.
2005;15:169–18.
Weacceptpre-submissioninquiriesOurselectortoolhelpsyoutondthemostrelevantjournalWeprovideroundtheclockcustomersupportConvenientonlinesubmissionThoroughpeerreviewInclusioninPubMedandallmajorindexingservicesMaximumvisibilityforyourresearchSubmityourmanuscriptatwww.
biomedcentral.
com/submitSubmityournextmanuscripttoBioMedCentralandwewillhelpyouateverystep:Nieetal.
BMCGenomics(2016)17:352Page16of16
- behavioriyunssr相关文档
- SCRiyunssr
- 19.iyunssr
- crucialiyunssr
- gliomaiyunssr
- Policyiyunssr
- 标记iyunssr
腾讯云2核4GB内存8M带宽 年74元
一般大厂都是通过首年才有可以享受爆款活动，然后吸引我们注册他们商家达到持续续费和购买的目的。一般只有大厂才能有这样的魄力和能力首年亏本，但是对于一般的公司和个人厂家确实难过，这几年确实看到不少的同类商家难以生存。这里我们可以看到有对应的套餐方案。不过这两个套餐都是100%CPU独享的，不是有某云商家限制CPU的。但是轻量服务器有个不好的就是带宽是较大且流量是限制的额，分别是1GB和1.2TB月流量...
Hostio€5/月KVM-2GB/25GB/5TB/荷兰机房
Hostio是一家成立于2006年的国外主机商，提供基于KVM架构的VPS主机，AMD EPYC CPU，NVMe硬盘，1-10Gbps带宽，最低月付5欧元起。商家采用自己的网络AS208258，宿主机采用2 x AMD Epyc 7452 32C/64T 2.3Ghz CPU，16*32GB内存，4个Samsung PM983 NVMe SSD，提供IPv4+IPv6。下面列出几款主机配置信息。...
Pacificrack：新增三款超级秒杀套餐/洛杉矶QN机房/1Gbps月流量1TB/年付仅7美刀
PacificRack最近促销上瘾了，活动频繁，接二连三的追加便宜VPS秒杀，PacificRack在 7月中下旬已经推出了五款秒杀VPS套餐，现在商家又新增了三款更便宜的特价套餐，年付低至7.2美元，这已经是本月第三波促销，带宽都是1Gbps。PacificRack 7月秒杀VPS整个系列都是PR-M，也就是魔方的后台管理。2G内存起步的支持Windows 7、10、Server 2003\20...
iyunssr为你推荐
-
网红名字被抢注球星名字被抢注合法合理吗？比肩工场比肩接踵的意思www.qq530.com谁能给我一个听歌的网站？777k7.comwww 地址 777rv怎么打不开了，还有好看的吗>com8090lu.com8090lu.com怎么样了?工程有进展吗？www.7788dy.com回家的诱惑 哪个网站更新的最快啊www.vtigu.com如图所示的RT三角形ABC中,角B=90°(初三二次根式）30 如图所示的RT三角形ABC中,角B=90°,点p从点B开始沿BA边以1厘米每秒的速度向A移动；同时，点Q也从点B开始沿BC边以2厘米每秒的速度向点C移动。问：几秒后三角形PBQ的面积为35平方厘米？PQ的距离是多少机器蜘蛛尼尔机械纪元机械蜘蛛怎么过 机械蜘蛛打法攻略解析www.dm8.cc有谁知道海贼王最新漫画网址是多少？？官人放题《墨竹题图诗》 大意