AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 133:10041012 (2007)

North Indian Muslims: Enclaves of Foreign DNA

or Hindu Converts?
Maria C. Terreros,1 Diane Rowold,1 Javier R. Luis,1,2 Faisal Khan,3 Suraksha Agrawal,3
and Rene J. Herrera1*
1

Department of Biological Sciences, Florida International University, University Park, Miami, FL 33199
Departamento de Genetica, Facultad de Biologa, Universidad de Vigo, Spain
3
Department of Medical Genetics, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow 226014, India
2

KEY WORDS

phylogenetics; mtDNA; North India; Y chromosome haplogroups

ABSTRACT
The mtDNA composition of two Muslim
sects from the northern Indian province of Uttar Pradesh, the Sunni and Shia, have been delineated using
sequence information from hypervariable regions 1 and
2 (HVI and HVII, respectively) as well as coding region
polymorphisms. A comparison of this data to that from
Middle Eastern, Central Asian, North East African,
and other Indian groups reveals that, at the mtDNA
haplogroup level, both of these Indo-Sunni and IndoShia populations are more similar to each other and
other Indian groups than to those from the other
regions. In addition, these two Muslim sects exhibit a
conspicuous absence of West Asian mtDNA haplogroups
suggesting that their maternal lineages are of Indian
origin. Furthermore, it is noteworthy that the maternal
lineage data indicates differences between the Sunni

and Shia collections of Uttar Pradesh with respect to

the relative distributions of Indian-specific M sub-haplogroups (Indo Shia > Indo Sunni) and the R haplogroup (Indo Sunni > Indo Shia), a disparity that does
not appear to be related to social status or geographic
regions within India. Finally, the mtDNA data integrated with the Y-chromosome results from an earlier
study, which indicated a major Indian genetic (Y-chromosomal) contribution as well, suggests a scenario of
Hindu to Islamic conversion in these two populations.
However, given the substantial level of the African/Middle Eastern YAP lineage in the Indo-Shia versus its absence in the Indo-Sunni, it is likely that this conversion
was somewhat gender biased in favor of females in
the Indo-Shia. Am J Phys Anthropol 133:10041012,
2007. V 2007 Wiley-Liss, Inc.

Present-day India is represented by a complex sociocultural mosaic comprised of 20 major languages and
*750 dialects (Kosambi, 1991) partitioned into *2,000
castes and tribal groups (Puppala, 1996). The vast majority of these ethnic populations (at least 80%) is Hindu
and socially organized into castes and subcastes (Karve,
1968). Tribal groups comprise about 8% of the total Indian population (Roychoudhury et al., 2001). A third
socio-religious group, the Muslims, which are represented by two major sects, the Sunni and Shia, constitute *12% of the total Indian population (Majumder,
2001).
The Sunni and Shia Muslim sects arose from a major
religious schism concerning the rightful lineage of the
Prophet Muhammads successor in the decades following his death (632 AD). The presence of these Islamic
settlements in India may be a result of at least three
distinct campaigns (Farah, 2003) initiated from different geographic regions. In 711 AD, an Arab military
invasion precipitated the formation of the Sind IndoMuslim state in the Indus delta region (Keay, 2000). A
few centuries later, between 997 and 1027 AD, Muslim
converts from the Central Asian Turkic tribe staged
multiple raids into the northwest province of Punjab.
Finally, during the 13th and early 14th centuries AD,
Afghan and Persian Muslims arrived from the northwest, reached New Delhi and from there, penetrated
into points east, west, and south (Wolpert, 1991). In
addition to different possible source populations, the
Indo-Muslim groups may have subsequently evolved
through several distinct cultural modes: cultural diffusion, elite dominance via military expansions, and colo-

nization, which may have involved varying levels of

genetic admixture with the indigenous Indian groups.
It has been stated that today most of the Muslims in
the Indian subcontinent represent the descendants of
converts and are the offspring of Hindu mothers (Wolpert, 1991). Alternately, the Indian Muslim groups may
be comprised almost entirely of descendents of Middle
Eastern or Central Asian migrants with minimal admixture from the surrounding populations. The present
study was designed to test these various hypotheses.
In this investigation, we hope to determine the relative degree of mtDNA affinity between two culturally
distinct Indian Muslim groups from the north central
province of Uttar Pradesh, the Indo-Sunni and IndoShia, and populations from four distinct geographic
regions: Middle East Asia, Central Asia, North East
Africa and the Indian subcontinent. In doing so, we aim
to delineate the mtDNA composition and the maternal
origin of these two Muslim enclaves. To accomplish this
objective, we compare mtDNA sequence information
from hypervariable regions one (HVI) and two (HVII) as
well as various coding polymorphisms.

C 2007
V

WILEY-LISS, INC.

*Correspondence to: Rene J. Herrera, Department of Biological

Sciences, Florida International University, University Park, Miami,
FL 33199. E-mail: herrerar@fiu.edu
Received 22 October 2006; accepted 29 January 2007
DOI 10.1002/ajpa.20600
Published online 11 April 2007 in Wiley InterScience
(www.interscience.wiley.com).

1005

GENETIC DIVERSITY IN MUSLIMS FROM NORTHERN INDIA

TABLE 1. Populations analyzed
Geographical area
Africa
North East
Asia
Middle East

West Central

Levant
Near East

South Central

Rest of Asia, India

North

East Central
West Central
South

Island

Country

Population

Language

Social
status

Reference

34
118

Egypt
Egypt

Gurna
Arabs/Berbers

Afro-Asiatic
Afro-Asiatic

Stevanovitch et al., 2003

Rowold et al., 2007

105
90
131
50
205

Oman
Qatar
UAE
Yemen
Central Asian

Afro-Asiatic
Afro-Asiatic
Afro-Asiatic
Afro-Asiatic
Altaic

Rowold et al., 2007

Rowold et al., 2007
Rowold et al., 2007
Rowold et al., 2007
Comas et al., 1998

388
192
139
39
69
117
50
451
42
37
20
17
216
41
44
100
44
38
33
44

Turkey
Armenia
Georgia
Jordan
Syria
Palestina
Arabia
Iran
Iran
Iran
Iran
Iran
Iraq
Turkmenistan
Tajikistan
Pakistan
Pakistan
Pakistan
Pakistan
Pakistan

Arabs
Arabs
Arabs
Arabs
Kirghiz, Kazakhs,
Uighurs
Anatolia
Armenian
Georgian
Arabs
Syrian
Palestinian
Arabs
Arabs
Persian
Gilaki
Kurdish
Lur
Iraqi
Turkmen
Shugnun
Karachi
Hunza
Brahui
Makrani
Parsi

Altaic
Indo-European
Caucasian
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Altaic
Indo-European
Indo-European
Isolate
Dravidic
Indo-European
Indo-European

Tambets et al., 2002

Tambets et al., 2002
Tambets et al., 2002
Rowold et al., 2007
Richards et al., 2003
Richards et al., 2003
Kivisild et al., 2003
Kivisild et al., 2003
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Al-Zahery et al., 2003
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,
Quintana-Murci et al.,

59
60
185
60
60
35
50
25
19
53
58
53
55
98

Uttar Pradesh

Shia
Sunni
Bhargava
Brahmin
Chaturvedi
Rajasthan
West Bengal
Rajbhansi
Bangladesh
Gujarat
Maharashtra
Karnataka
Kerala
Reddy
Thogataveera
Andra Pradesh
Sri Lanka
Tamil Nadu
Andaman

Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Indo-European
Dravidic
Dravidic
Dravidic
Indo-European
Indo-European
Indo-European
Dravidic
Andamanese

50
40
132

Rajasthan
West Bengal
Bangladesh
Gujarat
Maharashtra
Karnataka
Kerala
Andra Pradesh
Sri Lanka
Tamil Nadu
Andaman

Muslim
Muslim
Caste
Caste
Caste
Caste
Caste
Caste
Caste
Caste
Caste
Tribe
Caste
Caste
Tribe
Tribe
Caste
Tribe
Tribe

2004
2004
2004
2004
2004
2004
2004
2004
2004
2004
2004

Present study
Present study
Palanichamy et al., 2004
Palanichamy et al., 2004
Palanichamy et al., 2004
Metspalu et al., 2004
Metspalu et al., 2004
Palanichamy et al., 2004
Metspalu et al., 2004
Metspalu et al., 2004
Metspalu et al., 2004
Cordeaux etal., 2003
Metspalu et al., 2004
Palanichamy et al., 2004
Palanichamy et al., 2004
Cordeaux etal., 2003
Metspalu et al., 2004
Metspalu et al., 2004
Palanichamy et al., 2004

MATERIALS AND METHODS

Populations analyzed

tions was examined to allow comparative analyses

(Table 1).

A total 119 unrelated individuals were randomly

sampled from the Indo-Sunni and Indo-Shia (60 and
59, respectively), two Muslim communities from Uttar
Pradesh, a state in northern India. Three generation
pedigree charts were compiled to ensure the collection
of nonrelated subjects (Agrawal et al., 2005). Samples
were procured from different locales of Uttar Pradesh
(Agrawal et al., 2005). Additional mtDNA haplogroup
and control region sequence information from 44 published Indian, African and Eurasian reference popula-

Collection of samples
Blood samples were collected in EDTA Vacutainer
tubes. The nuclear fraction from peripheral leukocytes
was isolated as previously described (Luis et al., 2004).
Ethical guidelines were adhered to as stipulated by the
institutions involved in the research project. The individuals from the Indo-Sunni and Indo-Shia communities
gave their informed consent prior to their participation
in the study.

American Journal of Physical AnthropologyDOI 10.1002/ajpa

1006

M.C. TERREROS ET AL.

Haplogroup assignment

TABLE 2. mtDNA haplogroup frequencies

Genomic DNA was isolated from the peripheral leukocyte fraction of whole blood as previously described
(Agrawal et al., 2005). Both HVI and HVII regions were
PCR amplified using primers previously described
(Stoneking et al., 1991) and then sequenced and aligned
to the revised Cambridge reference sequence (rCRS)
(Anderson et al., 1981; Andrews et al., 1999).
Haplogroup classification is based on the sequence information from the mtDNA control region (HVI and nucleotide position 73 of HVII) and RFLP analyses as performed by Macaulay et al. (1999) assaying the restriction
status at the following sites: (4216 Nla III, 4830 Hae II,
5176 Alu I, 7025 Alu I, 9052 Hae II, 10397 Alu I, 10871
MnlI, 12308 Hinf I, 12629 Ava II, 12703 Mbo II, 14765
Acc I, 14766 Mse I). Additional studies (Kivisild et al.,
1999, 2003, 2004; Richards et al. 2000; Salas et al., 2002,
2004; Metspalau et al., 2004; Quintana-Murci et al.,
2004) were consulted to further delineate the haplotypes.
When feasible, the amplification protocol and primer sets
used to assay coding regions followed the procedure
described in Torroni et al. (1996). In some situations,
however, multiple bands resulting from the digestion of
monomorphic restriction sites present within the large
amplified fragment interfered with the determination of
restriction status on the agarose gel. Thus, for these
sites, we developed an assay involving internal primers
and much smaller amplicons (*200 nucleotides in
length). This strategy, referred to as reduced amplicon
RFLP, was very successful and resulted in easily discernable restriction digest patterns on the agarose gels. The
Baysean 0.95 credible regions (0.95 CR) are calculated
for each haplogroup by SAMPLING, a program kindly
provided by Vincent Macaulay.

Haplogroup frequencies

Statistical analyses
A principal component analysis (PC) [Numerical Taxonomy and Multivariate Analysis System or NTSYSpc2.02i by Rohlf (2002)] was performed using the haplogroup frequencies of the Indo-Sunni and Indo-Shia together with 44 previously studied populations from the
Middle East, Central Asia, North East Africa and India
(referenced in Table 1). Other statistical analyses
included the computation of standard diversity indices
and gene diversity scores (Nei, 1987) as well as an
EwensWatterson homozygosity test for selective neutrality (Ewens, 1972; Waterson, 1978). These procedures
were based on either the HVI alone (nps 1601916385)
or the HVI and HVII regions together (nps 16019
16385, nps 0007300400). In addition, an analysis of molecular variance (AMOVA) was conducted on haplogroup
frequency distributions of the extended data set including reference populations (Table 1) to assess population
genetic structure. These statistical procedures were executed using the Arlequin Version 2.0 package (Schneider
et al., 2000).
Reduced median networks of the relevant haplogroups
observed in the Indo-Sunni and Indo-Shia collections (M,
R, and U) were constructed by the NETWORK 4.1 software program developed by Fluxus Technology (www.
fluxus-engineering.com). Times to the most recent common ancestor (TMRCA) for mtDNA haplogroups were
estimated according to the methods of Forster et al.
(1996). As in previously published studies (Richards et
al., 2000; Salliard et al., 2000; Salas et al., 2002; Kivisild

Population
SHIA

SUNNI

Haplogroup
D
M3
M4
M5
M6
M9/M30
M25
M*
N*
R*
R5
R6
U
U7
X
M*
N*
R*
R6
U

0.95 credible
region
0.004
0.048
0.010
0.027
0.004
0.004
0.010
0.217
0.010
0.028
0.004
0.010
0.028
0.010
0.010
0.377
0.010
0.171
0.004
0.106

0.089
0.205
0.115
0.159
0.089
0.089
0.115
0.450
0.115
0.162
0.089
0.115
0.162
0.115
0.115
0.623
0.113
0.391
0.088
0.300

Frequencies
0.017
0.102
0.034
0.067
0.017
0.017
0.034
0.328
0.034
0.119
0.017
0.034
0.119
0.034
0.034
0.500
0.034
0.269
0.017
0.180

The M sub-haplogroups are based primarily on HVI motifs (np

16007 np 16381). M*, N*, R*, U* represent samples that cannot be resolved into downstream haplogroups. There is no overlap among haplogroup categories.

et al., 2004), the average mutation rate used is one nucleotide change in 20,180 years for nps 1609016365.

RESULTS
Haplogroup distribution
The resulting haplogroup frequencies along with the
Baysean 0.95 credible regions (0.95 CR) of the 119 individuals comprising the two Muslim Indian populations
are presented in Table 2 (individual haplotype information for the Indo-Sunni and Indo-Shia is provided in online
Tables 1a and 1b at http://http://www.fiu.edu/*herrerar/
Sunni_and_Shia_Haplotype_frequencies.xls). This table,
along with the phylogeographic map of Figure 1 (based
on the haplogroup frequencies of the two studied populations together with those of the reference studies), illustrate that the predominant mtDNA haplogroups in the
Indo-Sunni and Indo-Shia are M (0.500, 0.95 CR: 0.377
0.623 and 0.593 0.95 CR: 0.4650.709, respectively), R
(0.317, 0.95 CR: 0.2130.443 and 0.169, 0.95 CR: 0.095
0.285, respectively), and U (0.150, 0.95 CR: 0.0820.262
and 0.153, 0.95 CR: 0.0830.266, respectively). These frequencies are similar to that of the other Indian populations surveyed (Fig. 1). Especially noteworthy is the high
frequency of haplogroup M in India (0.549 overall) perhaps reflecting its ancient status within this subcontinent (Quintana-Murci et al., 1999). However, as seen in
Table 2, there are some differences in the M haplogroup
distribution between the Indo-Sunni and Indo-Shia. The
Indian-specific M sub-haplogroups of M3, M4, M5, M6,
and M25 (Metspalu et al., 2004) are present in the IndoShia [0.102 (0.95 CR: 0.0480.205), 0.034 (0.95 CR: 0.01
0.115), 0.067 (0.95 CR: 0.0270.159), 0.017 (0.95 CR:
0.0040.089), and 0.034 (0.95 CR: 0.010.115), respectively] but not in the Indo-Sunni.
Several uncommon mutations (16007; 16009; 16008;
16025; 16095) are detected (sometimes in combination

American Journal of Physical AnthropologyDOI 10.1002/ajpa

GENETIC DIVERSITY IN MUSLIMS FROM NORTHERN INDIA

1007

Fig. 1. Phylogeographic map of India (and other regions). This phylogeographic distribution is based on the geographical frequency distribution of the major mtDNA haplogroups (D, M, N, R, U, X and others) with N and M including only those samples not
subsumed by their respective sub-haplogroups (i.e. M: D, N: R, U and X, R: U).

with each other) in the Indo-Shia population. Three of

these, 16007, 16008, 16095, are present in both M and N
haplogroups which suggest that they predate the ancient
divergence of M and N. The other two, 16008 and 16025,
are observed only in N or M, respectively, perhaps reflecting a more recent origin. However, since there are 12 individuals in this study exhibiting these mutations, it is difficult to gauge the significance of this observed distribution.

Principal component analysis

The principal component analysis based on the relative frequencies of the mtDNA haplogroup categories of
the Sunni and Shia from Uttar Pradesh of North Central
India and 44 Middle East, Central Asia, North East
Africa as well as other Indian populations (from sources
referenced in Table 1) is shown in Figure 2. Axes I and

American Journal of Physical AnthropologyDOI 10.1002/ajpa

1008

M.C. TERREROS ET AL.

Fig. 2. Principal Component Analysis. Middle East: Persia (PER), Iran (IRA), Gilaki (GIL), Lur (LUR), Kurdish (KUR), Iraq
(IRQ), Arabia (ARA), Syria (SYR), Palestine (PAL), Anatolia (ANA), Jordan (JOR). Lower Arabian Peninsula: Oman (OMA),
Qatar (QAT), United Arab Emirates (UAE), Yemen (YEM). Central East Asia: Georgia (GEO), Central Asia (CAS), Turkmen,
Turkmenistan (TUR), Armenia (ARM), Shugnun, Tajikistan (SHU). India: Shia (SHI), Suni (SUN), Andaman (AND), Andhra Pradesh (APR), Bangladesh (BAN), Bhargava (BHA), Brahmin (BRA), Chaturvedi (CHA), Gujarat (GUJ), Karnataka (KAR), Kerala
(KER), Maharashtra (MAH), Rajasthan (RAJ), Rajbhansi (RAB), Reddy (RED), Sri Lanka (SRI), Tamil Nadu (TAM), Thogataveera
(THO), West Bengal (WBE). Pakistan: Karachi (KCH), Hunza (HUN), Makrani (MAK), Brahui (BRH), Parsi (PAR). North Africa:
Gurna (GUR), Egypt (EGY).

II account for 50.4% of the total variation (38.2 and

12.2%, respectively). Three distinct assemblies are apparent. The southern Arabian Peninsula, Levant, and
North East African populations are located in the upper
left quadrant. A tight cluster of mostly East Central
Asian, Near East Asian, and Middle Eastern groups is
contained in the lower left quadrant whereas the Indian
collections plot in the lower right hand corner of the
graph. Several East Central Asian and Pakistani groups
bridge the distance between the Indian cluster to the
right and the Middle East, Near East Asian, and Central
Asian groups to the left. The two Indo-Muslim populations segregate peripherally as part of the Indian assembly between the two upper caste groups from Uttar Pradesh (the Brahmin and the Bhargava) and the southwestern Indian groups of Kerala and Karnataka (caste
and tribal, respectively).

AMOVA, diversity indices, and neutrality test

The analysis of molecular variance (AMOVA) based on
the mtDNA haplogroup frequency data from 46 global
populations including the Sunni and Shia is presented in
Table 3. Significant genetic differentiation is detected
among the 11 geographic regions (North India, South

India, East Central India, Island, West Central India,

Near East Asia, Levant, Middle East, North East Africa,
South Central Asia, and West Central Asia) as well
as among populations within these regions (P-value
< 0.000001 for both tests) (Table 3). When partitioned
into seven groups by language, these 46 populations display significant mtDNA diversity (P-value 0.00782).
Significant differences are also observed among populations within these groups (P-value < 0.00001). There
appears to be a significant correlation between geographic partitioning and genetic variance among the Indian and Pakistani groups of (P-value < 0.00001). Also,
significant genetic differences (P-value 0.01271) are
observed among the four Indian social groups (upper
caste, lower caste, tribal, and Muslim). In contrast, however, there is no association between genetic variance
and linguistics among the Indian and Pakistani groups
(P-value 0.27370) (Table 3).
Population diversity values based on sequence information encompassing either the HVI alone (nps 16019
16385) or the HVI and HVII combined (nps 16019
16385; nps 0007300400) are listed in Table 4. The two
Indo-Muslim populations showed similar sequence diversity values (HVI: 0.989 and 0.992 for the Indo-Sunni:
and the Indo-Shia, respectively, HVI and HVII combined:

American Journal of Physical AnthropologyDOI 10.1002/ajpa

1009

GENETIC DIVERSITY IN MUSLIMS FROM NORTHERN INDIA

TABLE 3. AMOVA based on mtDNA haplogroup frequencies
% variation attributable to

Grouping

Fst

Geographical regions (all): 11 groups

Geographical regions (India): 5 groups
Linguistic groups (all): 7 groups
Linguistic groups (India): 3 groups
Social status (India): 4 groups

0.308
0.066
0.298
0.121
0.071

Among Groups
f
P-value
24.68
4.13
8.03
0.63
2.70

0.00000
0.00000
0.00782
0.27370
0.01271

Among populations
within groups
f
P-value

Within populations
f
P-value

4.22
3.43
21.80
11.56
4.46

71.10
92.44
70.18
87.81
92.84

0.00000
0.00000
0.00000
0.00000
0.00000

0.00000
0.00000
0.00000
0.00000
0.00000

For geographical regions and linguistic groups see Table 1.

Geographical (all): 11 groups [1-(North India) Indo-Sunni, Indo-Shia, Brahmin, Chaturvedi, Bhargava, Rajasthan; 2-(South India)
Thogataveera, Maharashtra, Andhra Pradesh, Sri Lanka, Tamil Nadu, Kerala, Karnataka, Reddy; 3-(East Central India) Bangladesh, West Bengal, Rajbhansi; 4-(Island) Andaman; 5-(West Central India) Gujarat; 6-(Near East) Persia, Iran, Gilaki, Lur, Kurdish, Iraq, Arabia, Shugnun; 7-(Levant) Syria, Palestine, Jordan; 8-(Middle East) Oman, Qatar, UAE, Yemen; 9-(North East Africa)
Gurna, Egypt; 10-(South Central Asia) Hunza, Karachi, Makrani, Branhui, Parsi; 11-(West Central Asia) Georgia, Central Asia,
Anatolia, Turkmen, Armenia]. Geographical (India): 5 groups [1-(North India) Indo-Sunni, Indo-Shia, Brahmin, Chaturvedi, Bhargava, Rajasthan; 2-(South India) Thogataveera, Maharashtra, Andhra, Sri Lanka, Tamil Nadu, Kerala, Karnataka, Reddy; 3-(East
Central India) Bangladesh, West Bengal, Rajbhansi; 4-(Island) Andaman; 5-(West Central Asia) Gujarat]. Linguistic (all): 7 groups
[1-(Indo-European) Indo-Sunni, Indo-Shia, Uttar Pradesh, Brahmin, Chaturvedi, Bhargava, Thogataveera, Maharashtra, Andhra
Pradesh, Bangladesh, West Bengal, Rajbhansi, Sri Lanka, Rajastan, Gujarat, Karachi, Makrani, Branhui, Parsi, Armenia, Shugnun, Persia, Iran, Gilaki, Lur, Kurdish, Iraq, Arabia, Syria, Palestine, Jordan; 2-(Dravidic) Tamil Nadu, Kerala, Karnataka, Reddy;
3-(Andamanese) Andaman; 4-(Afro-Asiatic) Gurna, Egypt, Oman, Qatar, UAE, Yemen; 5-(Isolate) Hunza; 6-(Caucasian) Georgia; 7(Altaic) Central Asia, Turkmen, Anatolia] Linguistic (India): 3 groups [1-(Indo-European) Indo-Sunni, Indo-Shia, Uttar Pradesh,
Brahmin, Chaturvedi, Bhargava, Thogataveera, Maharashtra, Andhra Pradesh, Bangladesh, West Bengal, Rajbhansi, Sri Lanka,
Rajastan, Gujarat; 2-(Dravidic) Tamil Nadu, Kerala, Karnataka, Reddy; 3-(Andamese) Andaman] Social status: 4 groups [1-(Muslim
sects) Indo-Sunni, Indo-Shia; 2-(Upper caste) Brahmin, Reddy; 3-(Lower caste) Andhra Pradesh, Uttar Pradesh, Tamil Nadu; 4(Tribal sects) West Bengal, Sri Lanka, Rajasthan, Maharashtra, Kerala, Karnataka, Gujarat, Bangladesh, Thogataveera, Chaturvedi, Bhargava, Rajbhansi].
TABLE 4. Diversity Indices and Neutrality Test
Sunni
No. of sequences
No. of haplotypes
Observed F value
Expected F value
Ewen-Watterson P value
Slatkins exact P value
Gene diversity 6 SD

Shia

HVI

HVII

HVI

HVII

50
41
0.0304
0.0291
0.9020
0.8980
0.9894 6 0.0071

60
57
0.0183
0.0184
0.8720
0.8720
0.9983 6 0.0034

60
51
0.0238
0.0227
0.9480
0.9440
0.9927 6 0.0052

55
48
0.0267
0.0235
0.9980
0.9980
0.9912 6 0.0070

0.998 and 0.991 for the Indo Sunni and the Indo-Shia,
respectively). Also, as shown in Table 4, the Indo-Muslim
populations yielded nonsignificant selection values for
the Ewen-Watterson neutrality test.

DISCUSSION
The Indo-Sunni and Indo-Shia enclaves in North Central India practice the Islamic religion and culture and
are thus, culturally distinct from the neighboring Indian
castes and tribal groups. However, the question remains
as to whether the Arab-based cultural elements of these
two North Indian Muslim groups are reflective of their
primary genetic affinity. In other words, do the Sunni
and Shia from North Central India share a higher
degree of mtDNA similarity with the groups from the
Middle East, where the Islam religion originated (7th
century AD in Saudi Arabia) and flourished? Alternatively, is the mtDNA composition closer to populations
from Central Asia (stronghold of the Mongol warlords
who had adopted the Islamic faith) or the surrounding
tribal and caste Indian groups? To answer these questions, mitochondrial DNA sequence information (HVI

TABLE 5. Times to the most recent common ancestor (TMRCA)

Haplogroup
M
R
U

All

Sunni

Shia

81.9 6 16.7
74.0 6 9.2
69.8 6 11.1

65.6 6 12.7
57.7 6 9.9
77.8 6 15.3

34.8 6 10.8
43.7 6 13.0
73.0 6 18.4

All TMRCA values are in thousands of years. M includes M*

and all other sub-types of M. R includes R*, R5 and R6, and U
encompasses U* and U7.

and HVII regions) and mtDNA coding polymorphism status was obtained for 119 unrelated Indo-Sunni (60) and
Indo-Shia (59) individuals.

mtDNA affinity: West Asian, Central Asian,

North East Africa, or Indian?
Our results demonstrate that the mtDNA haplogroup
M is present in the Indo-Shia and the Indo-Sunni Muslims at a frequency of 5060%, which is similar to that
found in Indian caste groups examined in this study and
others (Metspalu et al., 2004). Moreover, as seen in the
phylogeographic map (Fig. 1), there is a scarcity of the

American Journal of Physical AnthropologyDOI 10.1002/ajpa

1010

M.C. TERREROS ET AL.

M haplogroup in Western and Central Asia (Pakistan excluded). Noteworthy is the substantial overlap between
the TMRCAs of the Indo-Sunni (65,600 6 12,700 ybp)
(Table 5) and other Indian M lineages (53,000 6 7,000
ybp in Quintana-Murci et al. 1999) which indicates a
similar degree of M lineage diversity and age. Also,
these coalescent times are close to that of the initial out
of Africa dispersal of modern humans (54,000 6 8,000
ybp in Forster et al. (2001).
The somewhat younger TMRCA of the Indo-Shia M
haplogroup (34,800 6 10,800 ybp) (Table 5) may reflect a
more recent penetration into the Indian subcontinent.
Alternately, the younger TMRCA may also be due to a
lesser degree of admixture with the neighboring Indian
populations. However, the significant levels of Indian
specific types, M3, M4, M5, and M25 (0.237% 0.95 CR:
0.1470.360, v2 P-value < 0.0005) in the Indo-Shia versus their absence in the Indo-Sunni suggests otherwise.
A third possibility in which a recent episode of random
genetic drift reduced the variability of the M sub-haplogroups in the Indo-Shia population (thus, resulting in
a younger TMRCA) is discounted by the equally high
gene diversity indices of the Indo-Shia (0.9927 and
0.9912 for HVI and HVI and HVII combined, respectively) versus the Indo-Sunni (0.9894 and 0.9983 for HVI
and HVI and HVII combined, respectively).
The substantial R haplogroup levels in both the IndoSunni and the Indo-Shia (0.317%, 0.95 CR: 0.2130.443
and 0.169%, 0.95 CR: 0.0950.285, respectively) (Table
2) resemble those of the surrounding Indian groups (Fig.
1). In contrast, there appears to be a dearth of this haplogroup elsewhere (Africa, Europe, Central Asia, North
East Africa, and East Asia) except for some very low levels (
3%) in Oman, Persia, Qatar, UAE, and Yemen (AlZahery et al., 2003; Quintana-Murci et al., 2004; Rowold
et al., 2007).
The alignment of the Indo-Sunni and the Indo-Shia
with other Indian factions is also supported by the lack
of typical West Asian haplogroups such as H, I, J, K,
and T (Richards et al., 1998; Richards et al., 2000) in the
Indo-Muslims. Furthermore, the extremely low levels of
haplogroups U7 and W in the Indo-Muslims (two haplogroups common in Iran and Pakistan and/or polymorphic in the far northwest region of India) does not indicate substantial maternal admixture from North and
West Asia. These findings are also reflected in the segregation of populations in the PC plot (Fig. 2) and corroborated by the lack of specific lineage sharing (differing in
at least four mutations) between the Indo-Sunni and the
Indo-Shia Muslims on one hand and six Middle East
Asian or North East African populations on the other
(Jordan, Oman, Qatar, United Arab Emirates, Yemen
and Egypt; from Rowold et al., 2007).

mtDNA: Proto-Asian and caste affiliation

None of the Indo-Sunni and only one Indo-Shia individual (assigned to R5) possess any of the three ancient
mtDNA haplogroups (U2, M2, and R5) believed to be signatures of the first African to Eurasian exodus of modern humans (54,000 6 8,000 ybp in Forster et al., 2001;
70,000 6 10,000 ybp in Metspalu et al., 2004). These
original Indian types are represented in *15% of the
general Indian population and are not detected in other
parts of Eurasia (Metspalu et al., 2004). The apparent
absence of these ancient haplogroups in the two Muslim
collections of the current study may reflect a greater

degree of admixture with the more recent upper caste

migrants (Indo-European stock as listed in Table 1) in
the northern states (Metspalu et al., 2004).
Gene flow between northern upper castes and the
Indo-Sunni and Indo-Shia is also suggested by the Indian distribution pattern of the U haplogroup, a common
mtDNA component in East Central Asia and Europe
(Richards et al., 2000). This haplogroup is detected at
considerable levels in the two Indo-Muslim sects of Uttar
Pradesh (mean of 16.5%) and in the northern upper
caste populations nearby [a mean of 15.5% for Brahmin,
Bhargava, and Chaturvedi (Metspalu et al., 2004)] as
well as in the Hunza (34%) of northwestern India and
the Karachi population of Pakistan (17%) to the west
(Fig. 1). However, it is substantially reduced (a mean of
7.6%) in southern Indian populations, caste or tribal (Table 1 for source references).
The greater affinity of the Indo-Muslims with the
upper castes of northern India is logical in light of the
belief that high ranking Hindus may have descended
from European migrants 3,0008,000 ybp (Poliakov
1974; Renfew 1989a,b; Cavalli-Sforza et al., 1994; Bamshad et al. 2001). Moreover, it is possible that subsequent to the establishment of the caste system, the high
ranking Hindus were afforded a much greater opportunity to admix with western foreigners, which compounds
the genetic link between the Indo-Muslims and upper
Hindu castes.

Comparison between Indo-Shia and Indo-Sunni

The two studied groups are near each other within the
Indian clusters of the PC analysis (Fig. 2) indicating a
relatively close mtDNA haplogroup affinity to each other.
However, some differences exist such as the substantial
presence of Indian specific haplogroups (M3, M4, M5,
M6, M25, and R5) in the Indo-Shia (25.4%) versus its absence in the Indo-Sunni (v2 P-value < 0.0005). Also,
although the haplogroup R (any R that is not subsumed
by the haplogroups categories of U or X) is common to
both, it is significantly greater in the Indo-Sunni (v2 Pvalue < 0.025). It is unclear as to the significance of
these observations since, according to the results of Kivisild et al., 2003, the frequency patterns of the Indian
specific M sub-haplogroups or the R haplogroup do not
partition cleanly with respect to either geographic region
or social group of India.

Comparison with Y-chromosome data

Limited Y-chromosome data is available for these two
Indo-Muslim groups. The HG3 and HG9 frequencies of
the Indo-Muslim collection (unspecified as Sunni or Shia
in Rosser et al., 2000) (57.9% and 10.5% for HG3 and
HG9, respectively) are more similar to other North Indian groups than to the remaining Eurasian groups
examined in their study. The data from Rosser et al.
(2000) also underscores a very distinct partitioning of
the African/Middle Eastern YAP positive haplotype,
HG21. This lineage, also known as haplogroup E (Hammer,
2002), is defined by M145/M203, SRY4064 (Rosser et al.,
2000; Agrawal et al., 2005). It is present in respectable
frequencies in the Middle East and West Asia (mean of
16.1%) but is virtually absent in any of the North Indian
groups including the featured Indo-Muslim population
(Rosser et al., 2000). However, a later study (Agrawal
et al., 2005), distinguishes the Indo-Sunni from the Indo-

American Journal of Physical AnthropologyDOI 10.1002/ajpa

GENETIC DIVERSITY IN MUSLIMS FROM NORTHERN INDIA

Shia (the same two collections examined in the present
study) by substantially different levels of HG21. No
HG21 is detected in the Indo-Sunni as is the case for the
surrounding indigenous Indian groups. By contrast,
there is a substantial HG21 frequency in Indo-Shia
(11.3%), which resembles that of the Middle East collections of Agrawal et al. (2005) (mean of 16.3%) as well as
that of the Middle East and West Asian populations of
Mukherjee et al. (2001) and Rosser et al. (2000) (mean of
11.4% and 16.1%, respectively). An integration of these
findings with those of the current study suggests that
the Indo-Sunni and Indo-Shia are closer to each other
with respect to mtDNA haplogroup versus the Y-chromosome composition. In particular, although both IndoMuslim groups exhibit a high level of mtDNA admixture
with the indigenous Indian populations, the Indo-Shia
exhibits a substantially larger Middle East Y-chromosome component than the Indo-Sunni (which may have
been the Muslim sect featured in Rosser et al., 2000).
These findings may reflect the much greater assimilation
of Indian females versus males into the Indo-Shia sect.

SUMMARY
This study presents several notable findings. With
respect to mtDNA haplogroup distribution, both the
Sunni and Shia populations of Uttar Pradesh display a
much higher affinity to indigenous Indian populations
than to Middle Eastern and Central Asian groups. In
addition, the Indo-Shia displays a significantly higher
level of the Indian specific M sub-haplogroups (v 2 Pvalue < 0.001). The Indo-Sunni, on the other hand,
exhibits a significantly greater R frequency (v 2 P-value
< 0.025), which, like the Indian specific sub-haplogroups,
is also more common in India but not necessarily related
to social status (upper caste, lower caste, and tribal)
(Kivisild et al., 2003). In contrast, the available Y chromosome data suggests that the Indo-Shia population
exhibits higher Middle East Asian and lower North Indian Y chromosomal components compared with that of
the Indo-Sunni. Taken together, the mtDNA and Y chromosome data add support to the theory that the current
Sunni and Shia communities of Uttar Pradesh in northern India are mostly descended from Hindu converts.
For the Indo-Shia, however, the paternal haplogroup signature suggest a greater foreign Y chromosome contribution, possibly from the East or Central Asia, compared
with mtDNA. These results corroborate a scenario in
which mixed marriages in the past between Muslim
males and Hindu females engendered the majority of the
existent Muslim inhabitants.
Overall, our data suggests that the Y chromosome and
the mtDNA haplogroup composition of the Indo-Sunni
align this group closely to the neighboring Indian populations. MtDNA haplogroup differences between the
Indo-Shia and Indo-Sunni do not signal a differential
contribution of foreign mtDNA to either sect. Yet, they
may be indicative of unique population dynamics with
native castes and tribes.

LITERATURE CITED
Agrawal S, Khan F, Pandey A, Tripathi M, Herrera RJ. 2005.
YAP, signature of an African-Middle Eastern migration into
northern India. Curr Sci 88:19771980.
Al-Zahery N, Semino O, Benuzzi G, Magri C, Passarino G, Torroni A, Santachiara-Benerecetti AS. 2003. Y-chromosome and

1011

mtDNA polymorphisms in Iraq, a crossroad of the early

human dispersal and of post Neolithic migrations. Mol Phylogenet Evol 28:458472.
Anderson S, Bankier AT, Barrell BG, de Bruijn MH, Coulson
AR, Drouin J, Eperon IC, Nierlich DP, Roe BA, Sanger F,
Schreier PH, Smith AJ, Staden R, Young IG. 1981. Sequence
and organization of the human mitochondrial genome. Nature
290:457465.
Andrews RM, Kubacka I, Chinnery PF, Lightowlers RN, Turnbull DM, Howell N. 1999. Reanalysis and revision of the Cambridge reference sequence for human mitochondrial DNA. Nat
Genet 23:147.
Bamshad M, Kivisild T, Watkins WS, Dixon ME, Ricker CE,
Rao BB, Naidu JM, Prasad BV, Reddy PG, Rasanayagam A,
Papiha SS, Villems R, Redd AJ, Hammer MF, Nguyen SV, Carroll ML, Batzer MA, Jorde LB. 2001. Genetic evidence on the
origins of Indian caste populations. Genome Res 11:9941004.
Cavalli-Sforza LL, Piazza A, Menozzi P. 1994. The history and
geography of human genes. Princeton, NJ: Princeton University Press.
Comas D, Calafell F, Mateu E, Perez-Lezaun A, Bosch E, Martinez-Arias R, Clarimon J, Facchini F, Fiori G, Luiselli D, Pettener D, Bertranpetit J. 1998. Trading genes along the Silk
Road: mtDNA sequences and the origin of central Asian populations. Am J Hum Genet 63:18241838.
Cordeaux R, Saha N, Bentley G, Aunger R, Sirajuddin S,
Stoneking M. 2003. Mitochondrial DNA analysis reveals
diverse histories of tribal populations from India. Eur J Hum
Genet 3:253264.
Ewens WJ. 1972. The sampling theory of selectively neutral alleles. Theor Popul Biol 3:87112.
Farah CE. 2003. Islam. New York: Barons Educational Series.
Forster P, Harding R, Torroni A, Bandelt HJ. 1996. Origin and
evolution of Native American mtDNA variation: a reappraisal.
Am J Hum Genet 59:935945.
Forster P, Torroni A, Renfrew C, Rohl A. 2001. Phylogenetic
star contraction applied to Asian and Papuan mtDNA evolution. Mol Biol Evol 18:18641881.
Hammer M. 2002. A nomenclature system for the tree of human
Y-chromosomal binary haplogroups. Genome Res 12:339348.
Keay J. 2000. India. New York: Grove Press.
Karve I. 1961. Hindu society: an interpretation. Poona, India:
Deshmukh Prakashan.
Kivisild T, Bamshad MJ, Kaldma K, Metspalu M, Metspalu E,
Reidla M, Laos S, Parik J, Watkins WS, Dixon ME, Papiha
SS, Mastana SS, Mir MR, Ferak V, Villems R. 1999. Deep
common ancestry of Indian and western-Eurasian mitochondrial DNA lineages. Curr Biol 9:13311334.
Kivisild T, Reidla M, Metspalu E, Rosa A, Brehm A, Pennarun
E, Parik J, Gerberhiwot T, Usanga E, Villems R. 2004. Ethiopian mitochondrial DNA heritage: tracking gene flow across
and around the Gate of Tears. Am J Hum Genet 75:752770.
Kivisild T, Rootsi S, Metspalu M, Mastana S, Kaldma K, Parik
J, Metspalu E, Adojaan M, Tolk HV, Stepanov V, Golge M,
Usanga E, Papiha SS, Cinnioglu C, King R, Cavalli-Sforza L,
Underhill PA, Villems R. 2003. The genetic heritage of the
earliest settlers persists both in Indian tribal and caste populations. Am J Hum Genet 72:313332.
Kosambi DD. 1991. The culture and civilization of ancient India
in historical outline. New Delhi, India: Vikas Publishing House.
Luis JR, Rowold DJ, Regueiro M, Caeiro B, Cinnioglu C, Roseman
C, Underhill PA, Cavalli-Sforza LL, Herrera RJ. 2004. The Levant versus the Horn of Africa: evidence for bidirectional corridors of human migrations. Am J Hum Genet 74:532544.
Macaulay VA, Richards MB, Hickey E, Vega E, Cruciani F,
Guida V, Scozzari R, Bonne-Tamir B, Sykes B, Torroni A.
1999. The emerging tree of West Eurasian of West Eurasian
mtDNAs: a synthesis of control region sequences and RFLPs.
Am J Hum Genet 64:232249.
Majumder PP. 2001. Ethnic populations of India as seen from
an evolutionary perspective. J Biosci 26:533545.
Metspalu M, Kivisild T, Metspalu E, Parik J, Hudjashov G,
Kaldma K, Serk P, Karmin M, Behar DM, Gilbert MTP, Endicott P, Mastana S, Papiha SS, Skorecki K, Torroni A, Villems

American Journal of Physical AnthropologyDOI 10.1002/ajpa

1012

M.C. TERREROS ET AL.

R. 2004. Most of the extant mtDNA boundaries in South and

Southwest Asia were likely shaped during the initial settlement of Eurasia by anatomically modern humans. BMC
Genetics 5:125.
Mukherjee N, Nebel A, Oppenheim A, Majumder PP. 2001.
High resolution analysis of Y-chromosomal polymorphisms
reveals signatures of populations movements from central and
west Asia into India. J Genet 80:125135.
Nei M. 1987. Molecular evolutionary genetics. New York: Columbia University Press.
Palanichamy MG, Sun C, Agrawal S, Bandelt HJ, Kong QP,
Khan F, Wang CY, Chaudhuri TK, Palla V, Zhang YP. 2004.
Phylogeny of mitochondrial DNA macrohaplogroup N in
India, based on complete sequencing: implications for the peopling of South Asia. Am J Hum Genet 75:966978.
Poliakov L. 1974. The Aryan myth. New York: Basic Books.
Puppala S, Crawford MH. 1996. Genetic structure and phylogeny of populations of India: Andhra Pradesh. Homo 47:7384.
Quintana-Murci L, Chaix R, Wells RS, Behar DM, Sayar H,
Scozzari R, Rengo C, Al-Zahery N, Semino O, SantachiaraBenerecetti AS, Coppa A, Ayub Q, Mohyuddin A, Tyler-Smith
C, Mehdi SQ, Torroni A, McElreavey K. 2004. Where West
meets East: the complex mtDNA landscape of the southwest
and central Asian corridor. Am J Hum Genet 74:827845.
Quintana-Murci L, Semino O, Bandelt HJ, Passarino G, McElreavey K, Santachiara-Benerecetti AS. 1999. Genetic evidence
of an early exit of Homo sapiens sapiens from Africa through
eastern Africa. Nat Genet 23:437441.
Renfrew C. 1989a. Before Babel: speculations on the origins of
linguistic diversity. Camb Archaeol J 1:323.
Renfrew C. 1989b. The origins of Indo-European languages. Sci
Am 261:8290.
Richards MB, Macaulay V, Bandelt HJ, Sykes BC. 1998. Phylogeography of mitochondrial DNA in western Europe. Ann
Hum Genet 62:241260.
Richards M, Macaulay V, Hickey E, Vega E, Sykes B, Guida V,
Rengo C, Sellitto D, Cruciani F, Kivisild T, Villems R, Thomas
M, Rychkov S, Rychkov O, Rychkov Y, Golge M, Dimitrov D,
Hill E, Bradley E, Romano V, Cali F, Vona G, Demaine A,
Papiha S, Triantaphyllidis C, Stefanescu G, Hatina J, Belledi
M, Di Rienzo A, Novelletto A, Oppenheim A, Nrby S, AlZaheri N, Santachiara-Benerecetti S, Scozzari R, Torroni A,
and Bandelt HJ. 2000. Tracing European founder lineages in
the Eastern mtDNA pool. Am J Hum Genet 67:1251127.
Richards M, Rengo C, Cruciani F, Gratix F, Wilson JF, Scozzari
R, Macaulay V, Torroni A. 2003. Extensive female lineagemediated gene flow from sub-Saharan into Near Eastern Arab
populations. Am J Hum Genet 72:10581064.
Rohlf F. 2002. NTSYSpc. Setauket, NY: Exeter Publishing.
Rosser ZH, Zerjal T, Hurles ME, Adojaan M, Alavantic D,
Amorin A, Amos W, Armenteros M, Arroyo E, Barbujani G,
Beckman G, Beckman L, Bertranpetit J, Bosch E, Bradley
DG, Brede G, Cooper G, Corte-Real HB, de Knijff P, Decorte
R, Dubrova YE, Evgrafov O, Gilissen A, Glisic S, Golge M,
Hill EW, Jeziorowska A, Kalaydjieva L, Kayser M, Kivisild T,

Kravchenko SA, Krumina A, Kucinskas V, Lavinha J, Livshits

LA, Malaspina P, Maria S, McElreavey K, Meitinger TA,
Mikelsaar AV, Mitchell RJ, Nafa K, Nicholson J, Norby S,
Pandya A, Parik J, Patsalis PC, Pereira L, Peterlin B,
Pielberg G, Prata MJ, Previdere C, Roewer L, Rootsi S,
Rubinsztein DC, Saillard J, Santos FR, Stefanescu G, Sykes
BC, Tolun A, Villems R, Tyler-Smith C, Jobling MA. 2000.
Y-chromosomal diversity in Europe is clinal and is influrnced
primarily by geography, rather than by language. Am J Hum
Genet 67:15261543.
Rowold D, Luis JR, Terreros MC, Herrera RJ. 2007. Mitochondrial DNA gene flow indicates preferred usage of the Levant
corridor over the Horn of Africa passageway. Hum Genet
(in press).
Roychoudhury S, Roy S, Basu A, Banerjee R, Vishwanathan H,
Usha Rani MV, Sil SK, Mitra M, Majumder PP. 2001.
Genomic structures and population histories of linguistically
distinct tribal groups of India. Hum Genet 109:339350.
Saillard J, Forster P, Lynnerup N, Bandelt H-J, Nrby S. 2000.
mtDNA variation among Greenland Eskimos: the edge of the
Beringian expansion. Am J Hum Genet 67:718726.
Salas A, Richards M, De la Fe T, Lareu M-V, Sobrino B, Sanchiz-Diz P, Macaulay V, Carracedo A. 2002. The making of the
African mtDNA landscape. Am J Hum Genet 71:10821111.
Salas A, Richards M, Lareu M-V, Scozzari R, Coppa A, Torroni
A, Macaulay V, Carracedo A. 2004. The Africa diaspora: mitochondrial DNA and the Atlantic slave trade. Am J Hum
Genet 74:454465.
Schneider S, Roessli D, and Excoffier L. 2000. Arlequin ver
2000. A software for population genetics data analysis. Switzerland: University of Geneva.
Stevanovitch A, Gilles A, Bouzaid E, Kefi R, Paris F, Gayraud
RP, Spadoni JL, El-Chenawi F, Beraud-Colomb E. 2003. Mitochondrial DNA sequence diversity in a sedentary population
from Egypt. Ann Hum Genet 68:2339.
Stoneking M, Hedgecock D, Higuch1 RG, Vigilant L, Erlich HA.
1991. Population variation of human control region sequences
detected by enzymatic amplification and sequence specific oligonucleotide probes. Am J Hum Genet 48:370382.
Tambets K, Kivisild T, Metspalu E, Parik J, Kaldma K, Laos S,
Tolk HV, Golge M, Demirtas H, Geberhiwot T, Papiha SS, de
Stefano GF, Villems R. 2000. The topology of the maternal
lineages of the Anatolian and Trans-Caucasus populations
and the peopling of Europe: some preliminary considerations. In: Renfrew C, Boyle K, editors. Archaeogenetics:
DNA and the population prehistory of Europe. Cambridge:
McDonald Institute for Archaeological Research, p219
235.
Torroni A, Huoponen K, Francalacci P, Petrozzi M, Morelli L,
Scozzari R, Obinu D, Savontaus ML, Wallace DC. 1996. Classification of European mtDNAs from an analysis of three European populations. Genetics 144:18351850.
Watterson G. 1978. The homozygosity test of neutrality. Genetics 88:405417.
Wolpert S. 1991. India. Berkeley: University of California Press.

American Journal of Physical AnthropologyDOI 10.1002/ajpa