Molecular Insight into Caste Genesis

8/7/2019 Molecular Insight into Caste Genesis

1/41

Genome Biology2005, 6:P10

Deposited research articleMolecular insight into the genesis of ranked caste populations ofwestern India based upon polymorphisms across non-recombinantand recombinant regions in genomeSonali Gaikwad1 and VK Kashyap1,2*

Addresses: 1National DNA Analysis Center, Central Forensic Science Laboratory, Kolkata -700014, India. 2National Institute of Biologicals,Noida-201307, India.

Correspondence: VK Kasyap. E-mail: [email protected]

comme

nt

reviews

reports

depositedresearch

intera

ctions

information

refereedresearch

.deposited research

AS A SERVICE TO THE RESEARCH COMMUNITY, GENOME BIOLOGY PROVIDES A 'PREPRINT' DEPOSITORY

TO WHICH ANY ORIGINAL RESEARCH CAN BE SUBMITTED AND WHICH ALL INDIVIDUALS CAN ACCESS

FREE OF CHARGE. ANY ARTICLE CAN BE SUBMITTED BY AUTHORS, WHO HAVE SOLE RESPONSIBILITY FOR

THE ARTICLE'S CONTENT. THE ONLY SCREENING IS TO ENSURE RELEVANCE OF THE PREPRINT TO

GENOME BIOLOGY'S SCOPE AND TO AVOID ABUSIVE, LIBELLOUS OR INDECENT ARTICLES. ARTICLES IN THIS SECTION OF

THE JOURNAL HAVE NOT BEEN PEER-REVIEWED. EACH PREPRINT HAS A PERMANENT URL, BY WHICH IT CAN BE CITED.

RESEARCH SUBMITTED TO THE PREPRINT DEPOSITORY MAY BE SIMULTANEOUSLY OR SUBSEQUENTLY SUBMITTED TO

GENOME BIOLOGY OR ANY OTHER PUBLICATION FOR PEER REVIEW; THE ONLY REQUIREMENT IS AN EXPLICIT CITATION

OF, AND LINK TO, THE PREPRINT IN ANY VERSION OF THE ARTICLE THAT IS EVENTUALLY PUBLISHED. IF POSSIBLE, GENOME

BIOLOGY WILL PROVIDE A RECIPROCAL LINK FROM THE PREPRINT TO THE PUBLISHED ARTICLE.

Posted: 19 July 2005

Genome Biology2005, 6:P10

The electronic version of this article is the complete one and can befound online at http://genomebiology.com/2005/6/8/P10

2005 BioMed Central Ltd

Received: 18 July 2005

This is the first version of this article to be made available publicly and noother version is available at present.

This information has not been peer-reviewed. Responsibility for the findings rests solely with the author(s).


2/41


3/41

Abstract

Background

Large-scaletradeandculturalcontactsbetweencoastalpopulationsofwesternIndiaand

Western-Eurasianspavedforextensiveimmigrationandgenesisofwidespectrumof

admixedgenepool.TotraceadmixtureandgenesisofcastepopulationsofwesternIndia,we

haveexaminedpolymorphismsacrossnon-recombining20Y-SNPs,20Y-STRs,18mtDNA

diagnosticsites,HVS-1plusHVS-2regions;andrecombining15highlypolymorphic

autosomalSTRsinfourpredominantcastepopulations-upper-rankingDesasth-brahminand

Chitpavan-brahmin;amiddle-rankingKshtriyaMaratha;andalower-rankpeasantDhangar.

Results

Thegeneratedgenomicdatawascomparedwithputativeparentalpopulations-Central

Asians,WestAsiansandEuropeansusingAMOVA,PCplot,andadmixtureestimates.

Overall,disparateuniparentalancestries,andl.1%GSTvalueforbiparentalmarkersamong

fourstudiedcastepopulationslinkedwellwiththeirexchequerdemographichistories.

Marathi-speakingancientDesasth-brahminshowssubstantialadmixturefromCentralAsian

malesbutPaleolithicmaternalcomponentsupporttheirScytho-Dravidianorigin.Chitpavan-

brahmindemonstratesyoungermaternalcomponentandsubstantialpaternalgeneflowfrom

WestAsia,thusgivingcredencetotheirrecentIrano-ScythianancestryfromMediterranean

orTurkey,whichcorrelatedwellwithEuropean-lookingfeaturesofthiscaste.Thisalso

explainstheiruntraceableethno-historybefore1000years,brahminizationeventandlater

amalgamationbyMaratha.ThewidespreadPalaeolithicmtDNAhaplogroupsinMarathaand

DhangarhighlighttheirsharedProto-Asianancestries.MarathamalesharbouredAnatolian-

derivedJ2lineagecorroboratingtheblendingoffarmingcommunities.Dhangar

heterogeneityisascribabletopredominantlySouth-AsianmalesandWest-Eurasianfemales.


4/41

Conclusions

Thegenomicdata-setsofthisstudyprovideamplegenomicevidencesofdiverseoriginsof

fourrankedcastesandsynchronizationofcastestratificationwithasymmetricalgeneflows

fromIndo-EuropeanmigrationduringUpperPaleolithic,Neolithic,andlaterdates.However,

subsequentgeneflowsamongthesecasteslivingingeographicalproximity,havediminished

significantgeneticdifferentiationasindicatedbyAMOVAandstructure.


5/41

Background

MegadiversityinIndiaisascribabletodiversegenepoolsconstitutingProto-Australoid,

CaucasianMediterranean,Mongoliansandcompositeethnicstrains.Technological

innovationsoutsideIndiaresultedindemicdiffusionofNeolithicfarmersandmigrationof

Indo-European(IE)speakersontoIndiansub-continent[1].Archeologicalandlinguistic

evidencessupportthecommunalsocialsysteminIndo-Europeantribespracticing

AndronovocultureinCentralAsia.Theera4000-1500B.C.(Holoceneperiod)witnessed

thearrivalofIndo-Aryantribesequippedwithsuperiormilitarypower;spreadofnew

technologiesandIndo-Europeanlanguageamongthree-quarterofcontemporaryIndians[2,

3].Thechangedsocio-culturaldimensionsresultedintheculturalstratificationofIndo-

EuropeanandDravidianspeakersintoautochthonestribes(~450groups,8.08%)andHindu

castefold(80%)withmanyregionalsub-castes[1,4].ThecastesysteminIndiaconfigured

intofour-layerswasbaseduponprofessions:Brahminisapriestlyandlearnedupper-rank;

Kshatriyaiswarriorandaristocraticmiddle-rank;Vaishyaisalower-rankinglandownerand

tradingcaste;andShudraaresocialworkersformingthefourth-rank[5,6,7].The

Caucasoidinvaders:Greeks,Parthians,Scythians,andKushanswereassimilatedin

Kshatriyaclusterwhilethepre-existingsmallwarriorcommunitiesdidnotgetthecaste

status[8,9].Thecasteendogamyallowedsubsequentgenerationstobepriests,warriorand

businessmenofthesociety[10]andthissocialmechanismgeneratedanincrediblycomplex

genomicsub-structuringofIndiangenepool[11,12].Historically,northwestIndiafirstsaw

thepermanenthumansettlementsduetosustainablefoodproduction,storagepluscookingof

foodgrainsinpottery-ware.Thetradedevelopmentknittedthelocalsocialfabricintovillage

units[13].

Presentstudyexplorestheexistingsocialmosaic(castestructure)inwesternIndia

(MaharashtrasKonkancoastandnorthDeccanplateau)bygaininggenomicinsightsintothe


6/41

genesisoffourpredominant,rankedcastepopulationsincludingDesasth-brahmin,

Chitpavan-brahmin,Maratha,Dhangar(Table1).Theselectedpopulationsspeakeither

MarathiorKonkani-regionallanguagesbelongingtosouthernbranchofIndo-European

family(EthnologueWebSite).Thesocio-culturalenvironmenthadbeengreatlyinfluenced

bytheenroutecaravantradebetweennorthandsouthzone.Theflourishingtradecontacts

withthetechnologicallyadvanceWestern-EurasianculturessuchasGreece,Rome,andIran

resultedinexchangeoffoodspices,cotton,leather-merchandise,andpeacock-feathers

during1500B.C.-1200A.D.period[14]andpavedthewayforimmigrationofwhite

sailors/merchantsindifferenttimezonesandgrowthoffewIndo-Europeanfusiongroups.

Bene-Israelis(100-525A.D),Parsis(751A.D),Anglo-Indians(15th

century)andIndo-

Portuguese(14th

-15th

century)communitiesaresettledinwesterncoastalareas.

Previousstudiesbasedonlow-resolutionmarkersprovidedpreliminarydirectionintothe

understandingofgenepoolofwesternpopulations.Risleyin1915[15]conductedthefirst

everstudyonMaharashtrianBrahminsusing9anthropometricmarkersandspeculatedtheir

Scytho-Dravidianorigin.Later,KarveandMalhotra(1968)[16]implieddifferentorigins

ofMaharashtriancastesonthebasisofclassicalmarkers.Baigetal.(2004)[17]used

mitochondrialDNApolymorphismstoindicateearlylatePleistocenematernalrootsforboth

tribalandfewcastepopulationsofMaharashtra.However,theexchequerandvaried

migrationhistoriesofselectedfourpopulationsconferthemuniqueandsignificantfroma

geneticperspectiveandjustifyin-depthandextensivegenomeanalyses.Takingthese

assumptions,weperformedcomprehensiveanalysesonbatteryofsensitiveDNAmarkers

includingnon-recombiningpaternallytransmittedY-chromosomeandmaternallytransmitted

mtDNA;andrecombiningbiparentalautosomalSTRlociinfourselectedcastes.Innovative

andprecisescreeningtechniquesallowedrapidpopulationanalysesforautosomalgenetic

markers[6,18];mtDNAhaplotypesandhaplogroupaffiliations[19];andY-haplotypeson


7/41


8/41

(32%)inChitpavan-brahminfromKonkanascomparedto10%,meanSTRvariancewasestimatedtobe

highestinL(1.464),followedbyJ2(1.383),R1a(1.335),H(1.076),R2(1.002)andR1a1

(0.997).Among20Y-STRs,DYS426showed1or2allelesacrossallclades.Themodal

allele(mostfrequent)atDYS48was9onHbackgroundcomparedto11onR*andC

chromosomes.Theanalyzedsamplesweretwo-threestepmutationsawayfromCohen

ModelHaplotypeofNearEastJews.

MitochondrialDNAdiversity

Table4(seeadditionalfile2)represents75differentmtDNAhaplotypes(HVS-1andHVS-

2)in77individualsasonehaplotypeeachwassharedwithinMarathaandChitpavan-

brahminrespectively.TheHVS-1sequencemotifsandassociateddiagnosticmutations

clusteredinto25cladesandsub-cladesbelongingtomacrohaplogroupsM,N,andR.

SuperhaplogroupMispartitionedinto9sub-haplogroups;showshighestfrequency(64%)

followedbyR(18.7%),U(14.6%)andminorfractionofN(2.8%).MaharashtrainRclade

includeR*,H,HV,U*,K,J2;andNcompriseofN*,W.Phylogeneticrelationshipsamong

HVS-1haplotypesfallingintodifferenthaplogroupsarepresented(Figure1).


9/41

TheEarlyUpperPaleolithicSouth-AsianclusterMhasthehighestfrequencyfollowedby

M5,M4,M8c,andM25etc.Mlineagesweremostfrequent(72%)inMarathafollowedby

Chitpavan-brahmin(67%),Desasth-brahmin(53%)andDhangar(47%).TheotherSouth-

Asianspecificdiversesub-clustersofRandUwerealsofrequentinMaratha(89%),and

Chitpavan-brahmin(84%).U7andWlineagesassociatedwithanotherlateupperPaleolithic

migrationtoIndiansub-continentaccountedfor1%(CB)to10%(DB).Neolithicmigration

andveryrecentWesterninfluencewasseenmore(14-15%)inChitpavan-brahminand

Desasth-brahminthantheothertwocastes.ThestudiedcastesdidnotcarryA,B,M-Cand

M-Dhaplogroupsasindicatedby-663HaeIII,+5176AluI,+13259HincIIsitesand

absenceof9bpdeletionbetweenCOII/tRNALys

.

AsetofmtDNAdiversityindicesandtimeofdemographicexpansionforfourcastesis

showninTable5.Haplotypediversitieswerehighest(1.0000.017)inDhangarand

Desasth-brahminwhilehaplotypesweresharedwithinMarathaandChitpavan-brahmin.The

studiedgroupsdidnotsharehaplotypesamongthem.Thenucleotidediversitiesandmean

pair-wisedifferenceswerefoundtobehigherandsimilarinMarathaandDhangarthanthe

twobrahmincastes.Thefourcastesshowedunimodal(bell-shaped)mismatchdistributions

(figurenotshown).Inparallel,theraggednessindexwaslessthan0.05,negativevaluesof

FusFsandTajimasDdiffersignificantlyfromzero.Thesevaluesagreedwellwith

mismatchanalysisandprovidedclearevidenceofdemographicexpansionofcastesfrom

westernIndia.

Tauvaluebaseduponthelocationofthemismatchdistributioncrestprovidesarough

timeestimateofrapidexpansionofpopulation[43].Theestimatedvalueswerelargerthan

6.0correspondingtoexpansiontimesof>52,000YBPwhileChitpavan-brahminshowa

smallvalueof4.59conformingto39,000YBP.


10/41

AutosomalSTRanalyses

AllstudiedpopulationsexhibitednodetectabledeviationfromH-WExpectationwith

15loci.Thesegregatingalleleswerefoundtobehigherandsimilar(159)forMarathaand

Dhangar,intermediate(153)inDesasth-brahminandlower(142)inChitpavan-brahmin.DH

andCBsharedmostfrequentallelesat9lociandrareallelesrangingfrom0.6-0.7%at4

loci[26].

Singlelocusheterozygosity(Ho)estimatesrangedfrom0.612(CSF1PO;CB)to

0.948(PentaE;DH).Combinedheterozygositywasremarkablyhighandcomparableinfour

castes,rangingfrom77%(MA)to80%(DB).Averagevarianceinallelesizeconsidering15

lociwasestimatedtobeslightlyhigher(4.021)inChitpavan-brahmin,followedbyDhangar

(3.969),Maratha(3.920)andDesasth-brahmin(3.732).

GenomicvariationsbetweenMaharashtriancastes

TheGSTestimateatbiparentalmarkersvariedwidelyfrom0.002(D18S51)to0.041

(D7S820);combinedvalueforallthelociconsideredtogetherwas0.011indicatinglow

geneticdifferentiationbetweenpopulations.Usingstructureprogram,theproportionsof

individualsassignedtoeachclusterwereapproximatelythesamewithlittlevariation

betweenethnicgroupsundertheadmixturemodel.Thissymmetryisstronglysuggestiveof

theabsenceofpopulationstructureinthepresentstudy,sincerealpopulationstructureis

associatedwithindividualsbeingstronglyassignedtooneinferredclusteroranotherwiththe

proportionsassignedtoeachethnicgroupshowingasymmetry(Figure2)

PopulationclusteringasrevealedbyPCA

GeneticaffinitiesbetweenstudiedcastesandWestEurasians(seeTable1)were

testedinthelightofpopulationhistoriesusingPrincipalComponentanalyses.Resultsof

PCAarepresentedbyplotsofthefirsttwoPCs,whichtogetheraccountfor64%oftheY-


11/41

chromosomehaplogroup,68%ofmtDNAvariationand77%ofthebiparentalvariationsin

thesepopulations.

Figure(3a)representsthePCAofY-haplogroups,whereMaratha,Desasth-brahmin

andDhangarclustertogetherwhileChitpavan-brahminappearsasanoutlier,whichshow

geneticaffinitieswithWestAsians(Ashkenazi-Jews,Iranian);GreeksandCentralAsians

comparedtoverydistinctWest-EuropeanFrenchandPortuguese.PCplotwithbasalmtDNA

haplogroupfrequenciesasinputvectorsisshown(Figure3b),wherethefirsttwoPCaccount

for42.165%and25.724%ofthetotalvariationrespectively.ThefirstPCmainlyseparates

Desasth-brahminfromChitpavan-brahmin,thelattercasteoccupiedintermediateposition

betweenIndianandWestEurasian(Uzbeks,Turks,Iranian).ThepopulationsfromWestAsia

andEuropetendtoclustertogether.APCplotusingFSTdistancematrixbasedon7

autosomalSTRloci(Figure3c)showed77%varianceaccountingforfirsttwocomponents

indicatingasatisfactoryrepresentationoftheoriginaldata.Dhangar(DH)appearedasan

outlier.Chitpavan-brahmin(CB),Desasth-brahmin(DB)andMaratha(MA)clustertogether

andwereplacedclosertotheWest-Eurasianpopulations.

Evaluationofhierarchicalstructure,usingAMOVA

Recombiningandnon-recombiningDNApolymorphismsshowed>90%variations

withinpopulations(Table6).Thestudiedcastesshowedmaximumvariance(5.29%)atY-

hapolgroupsfollowedbymtDNAsequencediversity.Consideringsocio-hierarchy,the

upper,middle,andlower-rankingcastesshowedinsignificantvariancewhilestatistically

significantvariancewasapportionedbetweentwoBrahmincastesbasedontheirY-SNPand

mtDNAdata.MarathaandDhangarshowedgreatervarianceatY-haplogroupfollowedby

mtDNA.However,MarathiandKonkaniIndo-Europeanlanguagegroupsreveal

insignificantgeneticvariance,irrespectiveofgenomicdataset.Amoderatedifferencewas

observedbetweenthemtDNAsequencediversityofMarathi-speakers.Considering


12/41

AMOVAanalysesusinghaplogroupfrequencydataofstudiedcastes,CentralAsian,West

AsianandEuropeanpopulations,Y-haplogroupshowedmaximumgeneticvarianceamong

analyzedpopulations(9.14%)comparedtomtDNAlineages(4.36%).Thestudiedcastes

irrespectiveofrankaffiliationshowedlessergeneticvariancewithCentralAsiansandWest

AsianscomparedtoEuropeans.

WesternEurasiansadmixtureinwesterncastes

Admixtureestimatesbasedonthreesetsofgeneticmarkersarequantifiedand

summarized(Table7).Allfourcastesshowveryhigh(60-90%)indigenousgenetic

component.TheWesternEurasianadmixturerangedfrom10-40%.Male-specific

admixturefromMediterraneanbeltwashighest(4050%)inChitpavan-brahmin;from

CentralAsiaandEast-Europewas2228%inDesasth-brahmin;andfromCentralAsiaand

WestAsianwas20-25%inMaratha.MaternalgeneflowfromWestAsianswasprominent

inDhangarandDesasth-brahmin;EuropeanadmixturewasmoderateinChitpavan-brahmin.

Discussion

Comprehensiveanalysesofrecombiningbiparentalandnon-recombininguniparental

markershelpedinresolvingthescientificlacunaesurroundinggeneticstructure,affinities,

andoriginofrankedcastesofwesternIndia.EarlierstudiesonwesternIndianpopulations

werebasedonlow-discriminatinggeneticmarkers,whichprovidedinterestingpreliminary

informationontheirgeneticstructureandorigins[16,17,44].However,thecomposite

populationhistorycouldnotbedecipheredbecauseitlackstheveryessentialY-chromosomal

component.

Presentstudygaineddeepandfreshinsightsintothematernalandpaternalancestry

andbiparentalcompositionoffourhierarchicalcastepopulationspracticingdiversesocio-

culturaltraditionstotestthegeneralityofvariouscompetinghypothesisontheoriginsof

Indiancastepopulationsproposedinaforementionedstudies.Convincinggenomicevidences


13/41

forthediverseoriginsoffourwesterncastepopulationswassubstantiatedbythepresenceof

finerbranchesofmajormtDNAandY-chromosomehaplogroupsdatingbacktoearlyUpper

Paleolithic(30000-50000YBP)orrecentcoalescenceage(~10,000yearsago).Thehigher

proportionofmtDNAlineages(U2,W,R5,U7)infourcastessuggestedancestralgenepool.

However,West-EurasianparticularlyCentralAsianandEastEuropean-specificmale

lineages(R1a,R1a1,J2),andmtDNAlineages(H,K,HV,J2,U5,U3)werepresentat

variablefrequenciesinstudiedcastegroupsstronglyindicatingadmixtureandancestrywith

latestmigrants.Autosomalmicrosatellitediversityshowhighheterogeneity(78.4%)and

largenumber(622)ofsegregatingallelesacross15loci;howeverthethreesocialranksdid

notrevealsignificantgeneticvarianceandlowcoefficientofgenedifferentiation(GST=

1.1%)amongthemindicatedgeneticaffinitiesamongwesternpopulations.Theaboveresult

alsoimpliesrecentcommonoriginofsomegroupsandsubstantialgeneflowamongthe

similarrankingcastes.

ThegeneticaffinitywithWestern-Eurasiansisexplainedinthelightofimmigrationof

trademerchantstoKonkanwestcoastandconquestofScythians/Sakas,Kushansand

White-HunsoverIndo-GreekinIndusValleyduring1st

to5th

centuryB.C.[45,46];

subjugationofnativesinnorthwestIndia,formationofIndo-Scythianfusiongroups,

notablytheRajput,theMauryansandwesternKsatrapasinwesternIndia,who

overthrewSatavahansdynastyinDeccanplateau(Maharashtra)togaincontroloverthe

caravantraderoutesandalsosupportedBuddhistcavemonasteriesbythegeneratedwealth.

GenesisofBrahmincastes

Chitpavan-brahminandDesasth-brahminconstitutesjust10%ofentirepopulace(~80

million)ofwesternIndia.Theirdifferentmarriagerules,variedcustoms,differentlocal

dialects(southernbranchofIndo-Aryanlanguage)illustrateddistinctorigins.Ourextensive

comparativeanalysessupporttheirdifferentethno-histories.TheY-chromosomesofMarathi-


14/41

speakingDesasth-brahmincarriedR1a1lineageinhighfrequency,whichreflectedtheir

considerableaffinitywithCentralAsiangivingcredencetotheirScythicdescent

(admixture,PCplot,AMOVAanalysis).TheirintermediatemtDNAdiversitycompriseof

lowfrequencyWest-EurasiancladesandsignificantPaleolithicgenepool(M)indicating

South-Asianancestry,whichprovideevidenceoftheirtribaloriginduetoupwardsocial

mobilityoffemalesasshownbystudyofBaigetal.(2004).TheseBrahminsubjects

presentedhighestnumberofbiparentalalleles,heterozygosityandgeneticaffinitywith

CentralAsians.TheseanalysesprovideevidenceofScytho-DravidiangenesisofDesasth-

brahmin.Theyaretheancientupper-castecomprisingof50sub-divisionsorgotra[47]

becausetheconsiderabletime-depthasinferredfromTauvalue,helpedthemtoconsolidate

theirpredominanceindifferentadministrativejobsbesidestraditionalpriesthood.

Conversely,non-recombininguniparentalcontributionsinChitpavan-brahmin

MediterraneanorEastEuropeantypeasshownby20%(HV,U3)mtDNAlineagesand

highlyfrequent(R1aandL)Y-haplogroups.TheadmixtureandPCanalyses(Figure3a,b)

reflectedgeneticassociationofChitpavan-brahminwithIranian,Ashkenazi-Jews(Turkey),

Greeks(EastEurope)andtosomeextentwithCentralAsianTurkishpopulationselucidating

theirdistinctNordic,Scytho-Iranianancestry[48,49].TheCaucasianlinkofChitpavan-

brahminhasalsobeeninferredfrombiparentalmicrosatellitesvariations(Figure3c).The

observedgenomicanalysesassertedtheethnographicalfactthatChitpavan-brahminshare

ancestrywithconspicuouslyEuropean-lookingPaganorAlpinegroup,whounderreligious

pressurehadmigratedfromAnatolianTurkeyorEastEuropetoGujaratcoastprobablyvia

sea-vessel.Besides,theirdocumentedhistoryisuntraceablebeyond1000years,further

indicatingthattheywerenotpartoftheoriginalVedicmigrations(earlyIndo-European)on

thewestcoast.Therefore,thepresentgenomeanalysesprovideconclusiveevidenceoftheir

recentmigration,genesis,andexpansionaftertheymigratedfromSopara(Indiaswestern


15/41

tradezone)togeographicallyisolatedKonkan-region,wheretheyadoptedKonkani

language,andcultivatedcashcrop.TheirconsiderablegeneticaffinitywithMarathacaste

furthercorroboratedtheprevalentnormthatfewofthedynamicandintelligentChitpavans

wereBrahmanizedforperformingreligiousritualsinKingShivajiscourt(eliteMaratha

group)andsomemembersweregiventhetitleofPeshwaorMinisterformanagingthe

administrationofMarathakingdom,whichwasextendedfarthernorthafterKingsdeath

undertheirrule.Weobserved15%similarHVS-1sequencemotif(M4lineage)between

Chitpavan-brahminandBene-Israeli(orIndianJews),probablysuggestingsimilarindigenous

Paleolithiccontribution.ComparedtoDesasth-brahmin,Kokanasth-brahminshowedlowest

biparentaldiversity,youngerageofpopulationbaseduponTauvalue,largergeneticaffinity

withWestAsiansplusEastEuropeanssuggestingtheirrecentdescent,inabsenceof

bottleneckeffect.However,recentmarriagesbetweenDesasth-brahminboyandChitpavan

girlhavecontributedtowardstheirgeneticaffinityasshownstructureplot(Figure2).

OriginofpeasantryDhangarcaste

The23endogamoussub-castesofpastoralcasteclustershowconsiderablevariation

intheirnumericalstrengthandecologicaldistributions.Thecomplexethno-historysuggests

expansionandcontractionduringthelongevolutionaryhistoryofthispeasantcluster.Our

analysesonY-haplogroupsclearlyelucidatedProto-AsiangeneticancestryofDhangar,

whoseY-haplogroupdiversitywasslightlylowerthanMarathaevidentlysupportingfission

inDhangarclusterandfusioninMaratha.Contrastingly,highlevelsofWestAsianmaternal

componentinDhangarsuggestedasymmetricalgeneflow.The15biparentalmicrosatellite

markersshowedrichallelicdiversityandhighheterozygosityowingtoclusterexpansion.

Theprocessofgeneticfissionandfusionoflinguistically(Marathi,Hindi,Kannada,Telugu)

andoccupationallydiversegroupsisbetterunderstoodbyhighlevelsofdiversityin

Dhangarssub-castesofsouthernandnortheasternMaharashtra[50].Anearlierstudybased


16/41

onbilateralpalmarprintsin20sub-castes[51]provideexplanatoryevidencesonfissionof

singlecasteintosub-castes,namely,Hatkar,Zende,Thellari,andDange;andfusionof

linguisticallydifferentsubgroupssuchasAhirs,ShegarsfromnorthwestIndiaandKurmars

fromsouthIndiaforcarryingoutsimilaroccupationalpursuits.Thepracticeofinbreeding

occursinsomeisolatedgroupsformedasaresultoffission[51].Thus,theexistingsub-

structuringwithinaDhangarpeasantcasteisattributabletoinnumerablefissionandfusion

processesfordemographicandeconomicreasons.

GenesisofwarriorMarathacaste

Marathacasteisgeographicallydispersedandrepresentsmorethan50%ofthe

currentMaharashtrianpopulation.Thisendogamouscommunityoffifty-millionindividuals

becamedominantowingtotheiroccupationalhierarchy.Theirwarriorelementis

conglomerationofRoyaldescendentssuchasRashtrakuts,Mauryas,Pariharas/Parmar

(Pawar),Pratiharas,Shilahars,Kadambas,Yadavas,Chalukyasetc.asaresultofsuccessful

expeditionsandconquestsofdifferentparts(smallKingdoms)ofthecountry.Ouranalyses

showedlimitedfrequencyofHolocene-specificmtDNA(U5,HandW)buthigherfrequency

ofSouth-Asianlineages,substantiatingtheirPaleolithicancestry.Marathashowshigher

nucleotidediversity,mean-pairwisedifferencesthanBrahmincastesbutcomparablewith

Dhangarsuggestingcommonoriginsasaresultofgeneflowfrompeasantclusterasshown

bystructureplot(Figure2).ItalsodepictedaffinitywithtwoBrahmincaste,whichclearly

morethanexpectedmatrimonialalliancesbetweenupperandmiddle-rankingcastes[6,7].

Marathamales,inrelationtostudiedcastes,carriedhigherfrequencyofJ2lineage[22],

whichstronglyindicatedassimilationofAnatolianfarmersintoMarathagenepool.The

aboveobservationonceagainassertedthefactthatKshtriyagrouphasmixedgenepool

becauseofitsfluidgeneticboundary[3,45].


17/41

Conclusions

Ourcomprehensivegenomicanalysesshoweddivergentpaternalandmaternalancestryof

studiedfourcastescorrelatingwellwiththeirvariedmigrationandexchequerdemographic

histories.ThedistributionandadmixtureofWestern-Eurasian-specificmtDNAandY-

chromosomalhaplogroupslendsupporttothediversegenesisofwesternrankedcastes.The

asymmetricalProto-AsiancomponentandWestern-Eurasianadmixtureintwobrahmincastes

explainedtheScytho-Dravidianoriginofelite,ancientDesasth-brahminandmuchrecent

Irano-Scythianancestry(WestAsia,EastEurope)ofChitpavan-brahmin.Marathaand

DhangarhavesignificantPleistocenegenepoolcorroboratingtheirProto-Asianorigin.

MarathawarriorcastehasexperiencedgeneflowfromAnatolianagriculturist(J2)supporting

theconglomerationofmigrantagriculturalcommunities.TherecombiningSTRlocididnot

revealsignificantdifferenceinpopulationstructureattributingtohypergamybetween

BrahminsandMaratha,andsharedancestryofDhangarandMaratha.Thisstudy

interestinglysurmisesthesynchronizationofcastestratificationwithWest-Eurasians

admixtureinGangeticplains,whichspreadinwesternterritoryduetodemographicand

economicreasons.

Methods

ThePopulations

Inthisstudy,bloodspecimenswerecollectedinK3EDTAvialsfrom365unrelated,healthy

andconsentingindividualsbelongingtoDesasth-brahmin,Chitpavan-brahmin,Maratha,and

DhangarcastesinhabitingMaharashtra(20oNlatitude,76

oElongitude),westernIndia(see

MapsofIndiawebsite).ThestudyhasbeenundertakenwiththeapprovalofEthical

CommitteeofCFSL(Kolkata)andMHA,GovernmentofIndia.


18/41

Thesamplinglocations,relevantdemographicinformation,anddatasourcesregarding

casteandWestern-Eurasianpopulationsanalyzedfor3setsofmarkersaregiveninTable1.

AvailableY-andmtDNAhaplogroupdataofChitpavan-brahminandMaratha(seetable1

forspecificcitations)areincludedinourdataset.TheWestern-Eurasianpopulationsincluded

forcomparativeanalyseswereCentralAsians[Uzbek,Turkish,Kurd);WestAsians

[Ashkenazi-Jews,Arabs,Iranian];andEuropean[Portuguese,Greeks,French](seeTable1

forreferences).DuetopaucityofpublishedautosomalSTRdatasets,only7loci[vWA,

TH01,D18S51,D3S1358,D21S11,D8S1179,FGA]inTurkish,Ashkenazi-Jews,Iranian,

andthreeEuropeanpopulationswereconsideredtoexploregeneticaffinitiesandgeneflow

patterns.

SequencingandGenotypingstudy

GenomicDNAwasisolatedbystandardphenol-chloroformmethod[52].

Y-chromosometypingat20bi-allelicand20multi-allelicmarkerswasperformedin78

malesoutof120sampledmalesfrom4castes.Y-SNPs:M168,YAP,RPS4Y,M122,M89,

M172,M69,M9,92R7,M3,M207,M173,SRY1532,M17,M124,M18,M5,M20,M11,

andM70knowntoidentify12haplogroupsinEurasianpopulationsweretypedusing

validatedamplificationprotocols[20,53,54].Y-haplogroupnomenclaturewasdone

followingY-ChromosomeConsortium[55].

TheY-haplogroupswithrelatedhaplotypesaredistinguishableviamicrosatellitemarkers.A

singlemultiplexfluorescent-basedgenotypingassayof20Y-STRs(Table3,seeadditional

file1)wasdonefollowingprimerandamplificationconditions[56].STRampliconsoftwo

dinucleotide,16tetranucleotides,andtwopentanucleotides(table3forlocinames)were

separatedonABI3100GeneticAnalyzer,sizedbyGenescanTM

3.1andassignedallele

numberforcompleteY-haplotypeprofile.


19/41

Seventy-sevenmalesrepresenting4castesweresequencedatHVS-1(15997-16391)and

HVS-2(48-408)fragmentsusingBig-DyeTM

terminatorchemistry,purifiedbyethanol

precipitationandresolvedonABI3100GeneticAnalyzer(AppliedBiosystems).Super-

haplogroupsweredefinedbasedonRFLPdiagnosticmarkers[57].Thesub-haplogroupor

finerlineageswereassignedbyassayingadditionalinformativesites(Table4,additionalfile

2)byeithercompletemtDNAsequencing[58]ordiagnosticmarkersincoding-segments[19,

21,59].BothHVS-1motifandcoding-regionvariationswereusedtoclassifythematernal

lineagesaccordingtotheabovementionedpublishedsources.

All356sampledindividualsweregenotypedat15STRloci[twopentanucleotidesand13

tetranucleotides]co-amplifiedusingPowerPlex16multiplexkitandmanufacturers

instructions(PromegaCorp.,Madison,USA).

DataAnalyses

Y-STRhaplotypeswereconstructedforcastesamples(presentstudy).Y-STRvariancewas

estimatedfor6commonhaplogroups.mtDNAmutationswerescoredwithreferenceto

revisedCambridgeReferenceSequence[60].Reduced-MedianNetworkwasdrawnforHVS-

1haplotypesandtheirdiagnosticmarkersincastesubjects(ARohl;SharewarePhylogenetic

NetworkSoftwareWebSite).FifteenbiparentalSTRanalysesincludeestimationof

heterozygosity[61],allelesizevariance[62]andcoefficientofgenedifferentiation/GST[63]

basedonpublishedallelefrequenciesinfourcastepopulations[26].Thediversityindicesand

demographicparameters,viz.,Tauvalue,TajimasD,andFusFStestswereestimatedusing

Arlequinsoftware,version2.001[64].

TosubstantiatethehypothesisofcommonancestryofMarathaandDhangar,andgene

flowbetweentwoBrahmincastes,weanalyzedgenotypedataofunlinkedmarkersvia

admixture-modelofStructuresoftware,version2.0[65].Eachrunwasdoneafter100,000

burn-initerationsand1,000,000estimationiterationsforK=1to5


20/41

ThegeneticrelationshipsbetweenstudiedcastesandhistoricallyknownputativeWest-

Eurasianparentalpopulations(Table1)wereexaminedviaPrincipal-componentanalysis

(PCA)usingSPSS11.0package.Theinputvariableswereuniparentalhaplogroup

frequenciesandFSTdistancematrixfor7biparentalSTRloci.AMOVAapproachwasused

toestimateproportionofgeneticvarianceincasteandreferencepopulationsusing

ARLEQUINpackage.DetailedgroupingdesignsarelistedinTable6.

ADMIX95SoftwarebasedonGeneIdentitymethod[66]wasusedtoestimatethe

admixtureproportions(m+SE)ofWest-Eurasianpopulationsinthewesterncaste

populations.Theputativeparentalpopulationswereknowntohavehistoricaltradeand

culturallinkswithIndia.

Authors'contributions

SGperformedalllaboratoryexperiments,carriedoutstatisticalanalysesanddraftedthe

manuscript.VKKconceptualizedthehypothesisofthestudy,helpedincorrectinterpretation

ofresultsandimprovedthepresentationandstyleofthemanuscript.

Acknowledgements

AuthorsexpresstheirheartfeltgratitudetoallblooddonorsfromwesternIndia;Directorand

Dr.R.Trivedi,Chief,NationalDNAAnalysisCenter,CentralForensicScienceLaboratory,

Kolkataforprovidinglaboratoryfacilitiesandsupportinanalyses.Thepaperwouldnothave

beenpossiblewithoutextensiveinformationonethno-historicalbackgroundonstudiedcaste

populationsandintellectualinputsofDr.TSVasulu,IndianStatisticalInstitute,Kolkata.

ThefirstauthoracknowledgesstudentscholarshipfromDFS,MinistryofHomeAffairs,

Govt.ofIndia.


21/41

References

1. GadgilM,ShambuPrasadUV,ManoharanS,PatilS,JoshiNV:Peoplingof

India.In:TheIndianhumanheritage.Editedby:BalasubramanianD,Appaji

RaoN.Hyderabad:UniversitiesPress;1997:100-129.

2. RenfrewC:Archaeologyandlanguage:thepuzzleofIndo-Europeanorigins.

London:JonathanCape;1987.

3. Quintana-MurciL,KrauszC,ZerjalT,SayarSH,HammerMF,MehdiSQ,

AyubQ,QamarR,MohyuddinA,RadhakrishnaU,etal.:Y-chromosomeslineages

tracediffusionofpeopleandlanguagesinsouthwesternAsia.AmJHumGenet

2001,68:537-542.

4. MajumderPP:PeopleofIndia:Biologicaldiversityandaffinities.Evol

Anthropol1998,6:100-110.

5. KarveI:KinshiporganizationinIndia.Bombay:AsiaPublishingHouse;1968.

6. Cavalli-SforzaL,MenozziP,PiazzaA:Thehistoryandgeographyofhumangenes.

Princeton:PrincetonUniversityPress;1994.

7. SinghKS,ed:PeopleofIndia:Indiascommunities.VolumesIV-V.New

Delhi:OxfordUniversityPress;1998.

8. SpearP:AhistoryofIndia.Volume2.London:PenguinBooks;1975.

9. SastriKAN,SrinivasachariG:AdvancedHistoryofIndia.2ndedition.NewDelhi:

AlliedPublishersPvt.Ltd;1980.

10.MajumderPP:Indiancasteorigins:Genomicinsightsandfutureoutlook.

GenomeRes2001,.11:931-932.

11. BamshadM,KivisildT,WatkinsWS,DixonME,RickerCE,RaoBB,

NaiduJM,PrasadBV,ReddyPG,RasanayagamA,etal.:Geneticevidence


22/41

ontheoriginsofIndiancastepopulations.GenomeRes2001,11:994-1004.

12. RoychoudhuryS,RoyS,BasuA,BanerjeeR,VishwanathanH,UshaRani

MV,SilSK,MitraM,MajumderPP:Genomicstructuresandpopulation

historiesoflinguisticallydistincttribalgroupsofIndia.HumGenet

2001,109:339-350.

13. BhasinMK,WalterH:GeneticsofcastesandtribesofIndia.NewDelhi:Kamla-Raj

Enterprises;2001.

14. Klostermaier KK:A survey ofHinduism. 2nd edition. New York, Albany: SUNY

Press;1994.

15. RisleyHH:ThepeopleofIndia.Delhi:OrientBookReprintCorporation;1915.

16. KarveI,MalhotraKC:Abiologicalcomparisonofeightendogamous

groupsofthesamerank.CurrAnthropol1968,9:109-124.

17. BaigMM,KhanAA,KulkarniKM:MitochondrialDNAdiversityintribal

andcastegroupsofMaharashtra(India)anditsimplicationontheir

geneticorigins.AnnHumGenet2004,68:453-460.

18. ZhivotovskyLA,RosenbergNA,FeldmanMW:Featuresofevolutionand

expansionofmodernhumans,inferredfromgenomewidemicrosatellite

markers.AmJHumGenet2003,72:1171-1186.

19. MetspaluM,KivisildT,MetspaluE,ParikJ,HudjashovG,KaldmaK,Serk

P,KarminM,BeharDM,ThomasM,etal.:Mostoftheextant

mtDNAboundariesinSouthandSouthwestAsiawerelikelyshaped

duringtheinitialsettlementofEurasiabyanatomicallymodernhumans.

BMCGenet2004,5:26.

20. UnderhillPA,PassarinoG,LinAA,ShenP,MirzazonLahrM,FoleyRA,

OefnerPJ,Cavalli-SforzaLL:ThephylogeographyofY-chromosome


23/41

binaryhaplotypesandtheoriginsofmodernhumanpopulations.Ann

HumGenet2001,65:43-62.

21. Quintana-MurciL,ChaixR,SpencerWellsR,BeharDM,SayarH,Scozzari

R,RengoC,Al-ZaheryN,SeminoO,Santachiara-BenerecettiAS,etal.:

WhereWestmeetsEast:thecomplexmtDNAlandscapeofthesouthwest

andcentralAsiancorridor.AmJHumGenet2004,74:827-845.

22. RosserZH,ZerjalT,HurlesME,AdojaanM,AlavanticD,AmorimA,Amos

W,ArmenterosM,ArroyoE,BarbujaniG,etal.,:Y-chromosomaldiversity

inEuropeisclinalandisinfluencedprimarilybygeography,ratherthan

bylanguage.AmJHumGenet2000,67:1526-1543.

23. SeielstadMT,MinchE,Cavalli-SforzaLL:Geneticevidencesforahigherfemale

migrationrateinhumans.NatGenet1998,20:278-280.

24. CordauxR,SahaN,BentleyGR,AungerR,SirajuddinSM,StonekingM:

MitochondrialDNAanalysisrevealsdiversehistoriesoftribalpopulationsfrom

India.EurJHumGenet2003,11:253-264.

25. CordauxR,AungerR,BentleyGR,NasidzeI,SirajuddinSM,StonekingM:

IndependentoriginsofIndiancasteandtribalpaternallineages.CurrBiol2004,

14:231-235.

26. GaikwadS,KashyapVK:Polymorphismatfifteenhypervariablemicrosatellite

lociinfourpopulationsofMaharashtra,India.ForensicSciInt2002,126(3):

267-271.

27. Kivisild T, Rootsi M, Metspalu S, Mastana S, Kaldma K, Parik J, Metspalu E,

AdojaanM,TolkH-V,StepanovV,etal.: Thegeneticheritageofearliestsettlers

persistsbothinIndiantribalandcastepopulations .AmJHumGenet2003,72:

313-332.


24/41

28. AkbasakBS,BudowleB,ReederDJ,RedmanJ,KlineMC:Turkishpopulation

datawiththeCODISmultiplexshorttandemrepeatloci.ForensicSciInt2001,

123:227-229.

29. ZerjalT,SpencerWellsR,YuldashevaN,RuzibakievR,Tyler-SmithC:Agenetic

landscapereshapedbyrecentevents:Y-ChromosomalinsightsintoCentralAsia.

AmJHumGenet2002,71:466482.

30. PicornellA,TomasC,JimenezG,CastroJA,RamonMM:Jewishpopulation

geneticdatain20polymorphicloci.ForensicSciInt2002,125:52-58.

31. NebelA,FilonD,BrinkmannB,MajumderPP,FaermanM,OppenheimA:TheY

chromosomepoolofJewsaspartofthegeneticlandscapeoftheMiddleEast.

AmJHumGenet2001,69:10951112.

32. BeharDM,HammerMF,GarriganD,VillemsR,Bonne-TamirB,RichardsM,

GurwitzD,RosengartenD,KaplanM,PergolaSD,etal.:MtDNAevidencefora

geneticbottleneckintheearlyhistoryoftheAshkenaziJewishPopulation.Eur

JHumGenet2004,12(5):355-364.

33. Al-ZaheryN,SeminoO,BenuzziG,MagriC,PassarinoG,TorroniA,

Santachiara-BenerecettiAS:Y-chromosomeandmtDNApolymorphismsinIraq,

crossroadoftheearlyhumandispersalandofpost-Neolithicmigrations.Mol

PhylogenetEvol2003,28:458472.

34. KivisildT,RootsiS,MetspaluM,MetspaluE,ParikJ,KaldmaK,UsangaE,Mastana

S,PapihaS,VillemsR:TheGeneticsofLanguageandFarmingSpreadinIndia.

In:Examiningthefarming/languagedispersalhypothesis.Editedby:BellwoodP,

RenfrewC.Cambridge:TheMcDonaldInstituteforArchaeologicalResearch;2003:

215-222.

35. SantosSMM,BudowleB,SmerickJB,KeysKM,MorettiTR:Portuguese


25/41

populationdataonthesixshorttandemrepeatlociCSF1PO,TPOX,

TH01,D3S1358,VWAandFGA.ForensicSciInt1996,83:229-235.

36. AnjosMJ,CarvalhoM,AndradeL,Corte-RealF,VieiraDN,VideMC:Allele

frequenciesofSTRmultiplexsystemsintwoPortuguesepopulationsamples.

ProgrForensicGenet2000,8:208-211.

37. PereiraL,PrataMJ,AmorimA:DiversityofmtDNAlineagesinPortugal:nota

geneticedgeofEuropeanvariation.AnnHumGenet2000,64:491-506.

38. Sanchez-DizP,LareuMV,BrionM,SkitsaI,CarracedoA:STRdataforthe

AmpFlSTRProfilerPluslocifromGreece.ForensicSciInt2002,126:

265-266.

39. RichardsM,MacaulayV,HickeyE,VegaE,SykesB,GuidaV,RengoC,SellittoD,

CrucianiF,KivisildT,etal.:TracingEuropeanFounderLineagesintheNear

EasternmtDNApool.AmJHumGenet2000,67:1251-1276.

40. TofanelliS,TaglioliL,VaresiL,PaoliG:TheSTR-basedgeneticprofileof

thepopulationfromCorsicaisland(France).ForensicSciInt2001,123:

33-38.

41. RousseletF,PfitzingerH,ManginP:Multiplexamplificationand

automatedfluorescenttypingofshorttandemrepeat(STR)loci:The

FrenchExperience.AdvForensicHaemogenet1996,6:139-141.

42. RousseletF,PfitzingerH,ManginP:Apentaplexautomatedfluorescent

typingsystemforforensicidenficationandFrenchCaucasianpopulation

data.JForensicSci1997,42:500-503.

43.RogersAR,HarpendingH:Populationgrowthmakeswavesinthedistributionof

pairwisegeneticdifferences.MolBiolEvol1992,9:552-569.

44. RoyS,ThakurC,MajumderPP:MitochondrialDNAvariationinrankedcaste


26/41

groupsofMaharashtra(India)anditsimplicationongeneticrelationshipand

origins.AnnHumBiol2003,30(4):443-454.

45. ThaparR:AHistoryofIndia.VolumeI.London:PenguinBooks;1975.

46. KocharR:TheVedicpeople.Hyderabad:OrientLongman;2000.

47. SaojiAV,ed:SamasthMaharashtriyaBrahminPotshakha(Rootofall

MaharashtrianBrahmins,inMarathi).Nagpur:SukhadPrakashan;2000.

48. GuhaB:TheracialaffinitiesofthepeopleofIndia.In:CensusofIndia,1931,

PartIIIEthnographicalSimla:GovernmentofIndiaPress;1935.

49. DixitJ,SeriesEd:Chitpavanism,AtributetoChitpavanBrahminculture,

EncyclopediaMaharashtra.Volume1.Pune:DixitPublishers;2003.

50. MalhotraKC:InbreedingamongDhangarcastesofMaharashtra,India.JBiosoc

Sci1979,11:397-409.

51. KarmakarB:Distributionandpopulationvariationoftotalnumberofpalmar

triradiiamong20DhangarcastesofMaharashtra,India.Int.J.Anthropol1990,

5(3):255-269.

52. SambrookJ,FritschEF,ManiatisT:MolecularCloning:alaboratorymanual.

Editedby:FordN,NolanC,FergusonMC.ColdSpringHarbor:ColdSpringHarbor

LaboratoryPress;1989.

53. WhitfieldLS,SulstonJE,GoodfellowPN:SequencevariationofthehumanY-

chromosome.Nature1995,378:379-380.

54.UnderhillPA,JinL,LinAA,MehdiSQ,JenkinsT,VollrathD,DavisRW,Cavalli-

SforzaLL,OefnerPJ:DetectionofnumerousY-chromosomebiallelic

polymorphismsbydenaturinghigh-performanceliquidchromatography.

GenomeRes1997,7:996-1005.

55. Y-ChromosomeConsortium:Anomenclaturesystemforthetreeofhuman


27/41

Y-chromosomalhaplogroups.GenomeRes2002,12:339-348.

56. ButlerJM,SchoskeR,VallonePM,KlineMC,ReddAJ,HammerMF:A

novelmultiplexforsimultaneousamplificationof20YchromosomeSTR

markers.ForensicSciInt2002,129:1024.

57. RoychoudhuryS,RoyS,DeyB,ChakrabortyM,RoyM,RoyB,RameshA,

PrabhakaranN,UshaRaniMV,VishwanathanH,etal.:Fundamentalgenomic

unityofethnicIndiaisrevealedbyanalysisofmitochondrialDNA.CurrSci

2000,79(9):1182-1192.

58. TorroniA,RengoC,GuidaV,CrucianiF,SellittoD,CoppaA,CalderonFL,

SimionatiB,ValleG,RichardsM,MacaulayV,ScozzariR:Dothefourcladesof

themtDNAhaplogroupsL2evolveatdifferentrates?AmJHumGenet2001,

69(6):1348-1356.

59. PalanichamyMG,SunC,AgrawalS,BandeltHJ,KongQ,KhanF,WangC,

ChaudhuriTK,PallaV,ZhangY:PhylogenyofmitochondrialDNAhaplogroupN

inIndia,basedoncompletesequencing:implicationsforthepeoplingofSouth

Asia.AmJHumGenet2004,75(6):966-978.

60. AndrewsRM,KubackaI,ChinneryPF,LightowlersRN,TurnbullDM,HowellN:

ReanalysisandrevisionoftheCambridgereferencesequenceforhuman

mitochondrialDNA.NatGenet1999,23:147.

61. Nei M: Estimationofaverageheterozygosityandgeneticdistancefromasmall

numberofindividuals.Genetics1978,89:583-590.

62. Minch E, Ruiz-Linares A, Goldstein D, Feldman M, Cavalli-Sforza LL:

MICROSAT,Version1.0:acomputerprogramforcalculatingvariousstatistics

onmicrosatellitealleledata.In:StanfordUniversity,Stanford1997.


28/41

63. Nei M: Analysis of genediversity in subdividedpopulation.ProcNatlAcad Sci

USA1973,70(12):3321-3323.

64. Schneider S, Roessli D, Excoffier L: Arlequin, Version 2.0: a software for

populationgeneticdataanalysis.In:UniversityofGeneva,Geneva2000.

65. Pritchard JK, Stephens M, Donnelly P: Inference of population structure using

multilocusgenotypedata.Genetics2000,155:945-959.

66. ChakrabortyR:GeneIdentityinracialhybridsandestimationofadmixturerates.

In:Geneticmicrodifferentiationinmanandotheranimals.Editedby:NeelJV,Ahuja

YR.DelhiUniversity,Delhi:IndianAnthropologicalAssociation;1985:171-180.

Figures


29/41

Figure1.ReducedMedianNetworkrelating77mtDNAhaplotypesinwesternIndiancastesubjects.Circleareasareproportionaltohaplotypesfrequencies.Populationcodesare

asreportedintable1.Mandsub-clusters(bluecircles),Randsub-clusters

(pinkcircles),Uandsub-clusters(greencircles),Nandsub-lineages(redcircles)

Figure2.AssignmentofsamplesfromfourwesternIndiancastepopulationstogeneticClustersinferredfromtheSTRUCTUREanalysisforK=4.Figure3a.PrincipalComponentplotbasedonY-HaplogroupsfrequenciesamongcastesofwesternIndiaand9putativeparentalWestEurasiangroups.Table1for

populationabbreviation. Figure3b.PrincipalComponentplotbasedonmtDNA-haplogroupsfrequenciesamongcastesofwesternIndiaand9putativeparentalWestEurasiangroups.Table1

forpopulationcodes.Figure3c.PrincipalComponentplotbasedonbasedonFSTdistancematrixconstructedusingbiparentalSTRallelefrequenciesamongcastesofwesternIndiaand9putativeparentalWestEurasiangroups.Populationcodesasintable1.

Tables

Table1.Demographiccharacteristicsofstudiedpopulations,samplesizes,sourceof

genomicdataofstudiedandrelatedpopulation.

Table2.DistributionofY-haplogroupsincastepopulationsofWesternIndia.

Table5.mtDNAdiversityindicesanddemographicparametersincastepopulationsof

westernIndia.

Table6.AnalysesofMolecularVarianceincasteandputativeparentalpopulations

baseduponuniparentalandbiparentalmarkers.

Table7.Admixtureestimatesbasedonbi-anduni-parentalgeneticmarkersincaste

populationsofwesternIndia


30/41


31/41

Table1

Demographiccharacteristicsofstudiedpopulations,samplesizes,sourceofgenomicdataofstudiedandrelatedpopulations

A.Castes

Populations

[code]

Geographical

distributiona

Socialstatusandethnohistory Linguistic

familyb

AutosomalSTR(N) Y-haplogroup(N) mtDNAhaplogroup(

Desasth-brahmin[DB] WesternIndia Uppercaste,Indo-Caucasoidpool Marathi(IE) 102[26] 19(thisstudy) 19(thisstudy)

Chitpavan-brahmin[CB] WesternIndia Uppercaste,Nordicbuiltwithlight-coloreyes Konkani(IE) 67[26] 66=23(thisstudy)+

43[27]

77=20(thisstudy)+

57[27]

Dhangar[DH] WesternIndia Lowerpeasantrycastewith23sub-clusters;

geneticfusionorfissioncommon

Marathi(IE) 80[26] 17(thisstudy) 19(thisstudy)

Maratha[MA] WesternIndia Middlecastefordefensepursuits,blendofguerilla

groupsandagrarianclasses

Marathi(IE) 107[26] 19(thisstudy) 29=19(thisstudy)+

10[17]

B.Western-Eurasians

Uzbek[UZ] CentralAsia agriculturalism Altaic Notavaliable [29] [21]

Turkish[TK] CentralAsia pastoralnomadism AltaicTurkic [28] [29] [21]

Kurd[KT] CentralAsia agriculturalism IE Notavaliable [29] [21]

Ashkenazi-Jews[AJ] WestAsia,Israel Middleeasternancestry,geneticbottleneckby

mDNAstudy

Levantine

Arabic

[30] [31] [32]

Arabs[AR] WestAsian,Saudi

Arabia

ancestrywithnomadicSemitictribes,founderof

Islam,seafaringtradeinoilsector

GulfArabic Notavaliable [31] [33]

Iranian[IR] WestAsia,Iran AryanoriginfromIndo-Iraniantribes,founder-

Parsireligion

Farsi(IE) [30] [34] [19]

Portuguese[PT] WestEurope,

Portugal

technologicallyadvancebusinesscommunity Italic(IE) [35,36] [22] [37]

Greeks[GK] EastEurope,Greece AlexanderconqueredEurasia,Indo-Greekgroups

innorthwestIndia

Greek(IE) [38] [22] [39]

French[FR] WestEurope,France technologicallyadvancebusinesscommunity Italic(IE) [40,41,42] [22] [39]

aMaharashtramap(www.mapsofindia.com);

bIE:Indo-European,(www.ethnologue.com)

Genomicdatasourcesandsamplesizes(N)


32/41

Table2.DistributionofY-haplogroupsincastepopulationsofWesternIndia.

Y-Haplogroup Desasth-

brahmin

Chitpavan-

brahmin

Maratha Dhangar Average

Frequency

N=19 N=66 N=19 N=17

H 0.158 0.137 0.315 0.294 0.226

R1a1 0.368 0.045 0.211 0.235 0.215

R2 0.105 0.106 0.053 0.294 0.140

R1a 0.053 0.318 0.053 0 0.106

L 0.105 0.168 0.105 0.059 0.109

J2 0.105 0.121 0.21 0.059 0.124

C 0.053 0.03 0.053 0.059 0.049

K2 0.053 0.03 0 0 0.021

P* 0 0.03 0 0 0.008

F* 0 0.015 0 0 0.004

Haplogroup

diversity(SE)0.842(0.066) 0.834(0.026) 0.836(0.052) 0.809(0.055)


33/41

Table5.mtDNAdiversityindicesanddemographicparametersincastepopulationsof

westernIndia.

Maratha Dhangar Desasth-

brahmin

Chitpavan-

brahmin

N=19 N=19 N=19 N=20

Diversityindices

Polymorphicsites 70 61 57 53

Haplotypediversity(SE) 0.994(0.019) 1.000(0.017) 1.000(0.017) 0.995(0.017)

Nucleotidediversity(SE) 0.057(0.030) 0.050(0.027) 0.046(0.025) 0.045(0.023)

MeanPairwisedifference

()

11.491(5.451) 10.13(4.84) 9.368(4.502) 8.805(4.239)

Demographicexpansion

parameters

Fu'sFs -7.85(p=0.004) -11.24(p=0) -11.870(p=0) -10.424(p


34/41

Table6.AnalysesofMolecularVarianceincasteandputativeparentalpopulations

baseduponuniparentalandbiparentalmarkers

Groupings

mtDNA5,6 Y-SNP5 Y-STR Autosomal

STRmtDNAa,b Y-SNPa Y-STR Autosomal

STR

4Maharashtrian

castes

NA NA NA NA 1.71

(0.029)

5.29

(0.0009)

0.11

(0.240)

0.7(0.0)

3rankedcastes:

Brahmins,warrior

andpeasant

-1.76 -2.52 -0.1 0.47

(0.161)

3.18

(0.066)

7.16

(0.004)

0.19

(1.00)

0.29(0)

2Linguisticgroups:

Marathiand

Konkanispeakers

2.41

(0.25)

7.56

(0.27)

0.19

(0.268)

-0.33 0.60

(0.009)

-0.72 0.01

(0.344)

0.83(0)

4geographical

groups(castes,CA,

WA,EU)1,2,3

4.36(0.001)

9.14(0) NA NA 4.08(p=0) 6.03

(p=0)

NA NA

2(castes&CA) 5.8(0.034)

9.51(0.033)

NA NA 1.20(0.021)

4.51(0.001)

NA NA

2(castes&WA) 5.25(0.03)

9.58(0.03)

NA NA 5.28(0) 5.82(0) NA NA

2(castes&EU) 15.22(0.03)

15.52(0.03)

NA NA 0.68(0) 5.90(0) NA NA

1CA:CentralAsians,

2WA:

WestAsians,

3EU:Europeans;

5Kivisildetal.2003;

6Baigetal.2004

%variationattributableto

Amonggroups Amongpopulationswithingroups


35/41

Table7.Admixtureestimatesbasedonbi-anduni-parentalgeneticmarkersincaste

populationsofwesternIndia

CentralAsians1

WestAsians2

Europeans3

N=69 N=305 N=133

Chitpavan-brahmin:

Y-SNP 1 0.125 0.375

mt-DNA -0.081 -0.064 0.084(0.3)

autosomalSTR -0.0144 -0.084 0.209(0.06)

Desasth-brahmin:

Y-SNP 0.218 -0.062 0.28

mt-DNA 0.009(9.7) 0.158(4.52) -0.158

autosomalSTR -0.003 0.010(0.027) 0.212(0.020)

Maratha:

Y-SNP 0.273(2.7) 0.215 -0.273

mt-DNA 0.037(9.8) -0.146 -0.291

autosomalSTR -0.063 0.42(0.009) 0.03(0.01)

Dhangar:

Y-SNP -0.625 0.065(9.1) 0.048

mt-DNA 0.011 0.238 -0.216

autosomalSTR 0.054(0.005) 0.075(0.001) -0.47(0.008)

*Table1forparentalcomposition1Turkish;

2Ashkenazi-Jews,Iranian;

3PT,GK,FRforautosomalSTRloci

1.34(0.006)

0.95

0.97

0.78(6.5)

1.4(7.6)

0.994(0.019)

0.56

0.972(0)

0.78(0.03)

1.062(9.16)

0.89(0.041)

%contribution[m(SE)]fromputativeparentalpopulations*

extantMaharashtrian

castes

-0.500

N=55-102


36/41


37/41

igure 2


38/41


39/41

Figure 4


40/41

Figure 5


41/41

Additionalfilesprovidedwiththissubmission:Additionalfile2:Table4-additionalfile2.xls:169Kb

http://genomebiology.com/imedia/1995289715740764/sup2.XLSAdditionalfile1:Table3-additionalfile1.xls:42Kbhttp://genomebiology.com/imedia/1328867408740764/sup1.XLS

Documents

Molecular Insight into Caste Genesis