Microsatellite Diversity Delineates Genetic Relationships of Shia and Sunni Muslim Populations of Uttar Pradesh, India

Muthukrishnan Eaaswarkhanth Bhawna Dubey Poorlin Ramakodi Meganathan Sabahat Noor

Human Biology, Volume 81, Number 4, August 2009, pp. 427-445 (Article)

Published by Wayne State University Press DOI: 10.1353/hub.0.0071

For additional information about this article

http://muse.jhu.edu/journals/hub/summary/v081/81.4.eaaswarkhanth.html

Access Provided by Goteborgs universitet at 07/10/12 7:31PM GMT

Microsatellite Diversity Delineates Genetic Relationships of Shia and Sunni Muslim Populations of Uttar Pradesh, India
Muthukrishnan Eaaswarkhanth,1 Bhawna Dubey,1 Poorlin Ramakodi Meganathan,1 Sabahat Noor,1 and Ikramul Haque1
Abstract In this study we characterize the genetic diversity and relationships between the Shia and Sunni Muslim populations of North India and geographically targeted neighboring and global populations. We examined a number of parameters of population genetic and forensic interest based on the allele frequencies from 15 autosomal STR loci (D8S1179, D21S11, D7S820, CSF1PO, D19S433, VWA, TPOX, D18S51, D3S1358, THO1, D13S317, D16S539, D2S1338, D5S818, and FGA). All the studied loci were consistent with Hardy-Weinberg equilibrium, except loci D18S51 and FGA for both Muslim populations, even after applying the Bonferroni correction. The combined power of exclusion and combined power of discrimination values for all 15 STR loci were 0.9999 and >0.99999, respectively, in both Muslim populations. Gene diversity values ranged from 0.6784 (TPOX) to 0.9027 (FGA) for Shia Muslims and from 0.7152 (CSF1PO) to 0.9120 (D18S51) for Sunni Muslims. The observed heterozygosity (Ho) ranged from 0.5833 (D18S51) to 0.8595 (VWA) in Shia Muslims and from 0.6818 (CSF1PO) to 0.8333 (D21S11) in Sunni Muslims and was lower than the expected heterozygosity (He) for 11 out of the 15 STRs typed. We analyzed the genetic affinities of the Shia and Sunni Muslim populations with their geographically closest neighboring North Indian, Middle Eastern, East Asian, and European populations using distance-based methods, including neighbor-joining trees and multidimensional scaling. In addition, we estimated the genetic contribution of the putative parental populations included in the analysis to the Shia and Sunni Muslim gene pool using admixture analysis. Although we observed a certain degree of genetic contribution from Iran to both Muslim populations, the results of the phylogenetic analyses based on autosomal STRs suggest genetic relatedness with some of the geographically closest neighboring Hindu religious populations.

1 National DNA Analysis Center, Central Forensic Science Laboratory, 30 Gorachand Road, Kolkata 700 014, West Bengal, India.

Human Biology, August 2009, v. 81, no. 4, pp. 427445. Copyright 2009 Wayne State University Press, Detroit, Michigan 48201-1309

Key words: population genetics, AmpFlSTR Identifiler, short tandem repeats (STRs), Shia Muslims, Sunni Muslims, genetic affinities, D8S1179,
D21S11, D7S820, CSF1PO, D19S433, VWA, TPOX, D18S51, D3S1358, THO1, D13S317, D16S539, D2S1338, D5S818, FGA, amelogenin.

428 / eaaswarkhanth et al. India is the largest secular country with a polygenetic population, and it is said to be the melting pot of various ethnic groups. Among the various known religions found in India (Islam, Christianity, Sikhism, Buddhism, Jainism, and Judaism), Islam is the second-most practiced religion, after Hinduism. Muslims of India make up more than 13% of the population (Census of India 2001), and they belong to various linguistic and ethnic groups of different biradaris (so-called castes) and a few tribes. Islamic presence was first felt in the Indian subcontinent in the 7th century, when Arab military forces conquered Sindh and added it to the Arabian Empire (Robb 2002; Schimmel 1982). Then, Sindh was established as an Indo-Islamic state, and later the region served as an Islamic outpost where Arabs established their trade links with the Middle East. At the end of the 10th century, remarkable changes took place when the Central Asian Turkic tribes embraced and took up the mission of spreading Islam. Eventually, a Turkic kingdom was established in Delhi during the 13th century, which facilitated Persian and Afghan Muslim invaders to further spread across India. By the 16th century, India was mainly ruled by Muslims. Muslim sultanates ruling from Delhi (beginning in the 11th century) and the great Mughal Empire that followed (15261857) created a substantive Islamic legacy before India fell under British colonial rule (Robb 2002; Schimmel 1982). The emperors of the Mughal dynasty were Sunnis for the most part. After the departure of the Mughal emperor Aurangzeb in 1707, his successors were more tolerant and even leaned heavily toward Shiism. The greater tolerance of the Mughal Empire allowed more elite recruitment of avowed Shias from Iran to high office. These Iranian Shias, the nawabs, originally temporary Mughal appointees, gained the confidence and support of the Hindu peasants and Sunni townsmen of Awadh. This situation favored three nawabs to form a strong Shia state in Awadh during the period 17221775. Eventually a Shia dynasty flourished, with the capital at Lucknow, and lasted for 136 years (Hollister 1988). Lucknow is the capital of modern-day Uttar Pradesh. Uttar Pradesh is an ancient land covering a large part of the highly fertile and densely populated upper Gangetic plain. It is recognized in the later Vedic age as Madhya Desha (midland) or Aryavarta (the Aryan land) (Lal 2005). It is the fifth largest state of India, with an area of 243,286 km2 and a population of 167 million (Census of India 2001). The state shares an international border with Nepal to the north and is surrounded by the states of Uttarakhand, Himachal Pradesh, Haryana, Delhi, Rajasthan, Madhya Pradesh, Chhattisgarh, Jharkhand, and Bihar to the east, south, and west (Figure 1). Because of Islamic influence for centuries, Uttar Pradesh became the heartland of Indian Muslims, becoming home to 32 million Muslims, nearly one-third of Indias total Muslim population (Census of India 2001). Classical genetic marker studies have shown that the genetic diversity of Uttar Pradesh Muslims takes two forms: (1) genetic affinity of the Muslim groups with their neighboring Hindu caste populations, suggesting that they are the descendants of converts from the seventeenth century who are still at an early stage

Genetic Relations of Shia and Sunni Muslims / 429

Figure 1.

Uttar Pradesh and the geographic location of the Muslim populations included in the present study.

of differentiation (Lanchbury et al. 1996: 655; Papiha 1996); and (2) genetic differentiation of some Muslim communities from the indigenous Hindu populations, perhaps resulting from a higher proportion of genetic lineages of external origin (Aarzoo and Afzal 2005; Papiha 1996). Interestingly, two points should be noted. First, the greater affinity between Muslims of Delhi and Muslims of Uttar Pradesh indicates that Delhi and Lucknow, the capitals during the Mughal Empire, provided greater opportunity for invaders to contribute to the gene pool of the regional populations (Papiha 1996). Second, differentiation among the few Muslim communities from western Uttar Pradesh suggests a different origin and ancestry for these groups (Ara et al. 2008) and reveals a difference among the Muslim communities of northern India. Thus classical marker studies are in agreement with the historical and anthropological fact that contemporary Indian Muslims are either the descendants of local converts or the descendants of Muslim invaders. Our recent reports based on 13 autosomal STR markers have shown the existence of regional affinity between one South Indian Muslim population and its geographically closest neighboring Hindu religious populations (Eaaswarkhanth et al. 2008, 2009). Therefore a similar trend of regional affinity rather than religious affinity can be expected to exist among North Indian Muslims as well. Against the background of the classical marker data (Ara et al. 2008; Papiha 1996) and historical information (Hollister 1988; Robb 2002; Schimmel 1982), we analyzed the genetic diversity patterns and affinities of Uttar Pradesh Muslims by categorizing them according to their sect: Shia and Sunni. Because of the lack of population data for the entire set of 15 STRs, we therefore used the commonly available 13 CODIS core STR loci from the worldwide data set. With this subset of 13 STR markers, we analyzed relationships among Shia and Sunni Muslim

430 / eaaswarkhanth et al. populations of North India and geographically targeted populations from northern India, Europe, the Middle East, and East Asia.

Materials and Methods

Subjects. Blood samples were obtained from 253 unrelated healthy individuals from Lucknow (Shia Muslims, n = 121; and Sunni Muslims, n = 132) after informed consent was acquired. We followed the ethical guidelines stipulated by the ethical review committee of the Central Forensic Science Laboratory, Kolkata, India. To determine population affinities and genetic admixtures, we compared the STR diversity of Shia and Sunni Muslims with the published allele-frequency data of geographically targeted population groups (listed in Table 1) based on the subset of 13 autosomal STR markers (common in the set of 15 STR markers). DNA Extraction and STR Typing. Genomic DNA was extracted using the standard phenol-chloroform procedure (Sambrook et al. 1989) and was purified using ethanol precipitation. Multiplex PCR amplification was performed for 16 STR loci using the AmpFlSTR Identifiler PCR Amplification Kit according to the user manual specifications (Applied Biosystems 2001). The 16 loci amplified simultaneously in this study were D8S1179, D21S11, D7S820, CSF1PO, D19S433, VWA, TPOX, D18S51, D3S1358, THO1, D13S317, D16S539, D2S1338, D5S818, FGA, and the gender determination marker amelogenin. PCR products were genotyped using an ABI 3100 automated DNA sequencer (Applied Biosystems, Foster City, California). Gene Scan analysis software, version 3.7, was used to determine the allele fragment size. Genotyper software, version 3.7 for Windows NT, was used to designate alleles by comparison with the allelic ladder supplied with the kit. Statistical Analyses. Allele frequencies of the 15 STR loci were calculated using GenePop, version 3.4 (Raymond and Rousset 1995). The Arlequin software package, version 3.1 (Excoffier et al. 2005), was used to determine HardyWeinberg p values, observed heterozygosity (Ho), expected heterozygosity (He), and gene diversity values. Locus-wise gene diversity (GD) and polymorphic information content (PIC) were computed using Power Marker, version 3.25 (Lui and Muse 2005). Several parameters of population genetic importance, such as matching probability ( pM), power of discrimination (PD), power of exclusion (PE), and paternity index (PI), were calculated using the Excel PowerStats spreadsheet (www.promega.com/geneticidtools/powerstats). The allele frequencies of the 13 STR markers for the compiled populations (see Table 1) were used to generate neighbor-joining (NJ) trees based on Neis DA genetic distances using various options in PHYLIP 3.6 (Felsenstein 1993). Bootstrap consensus scores (1,000 replications) were generated using the SEQBOOT and GENDIST options of the PHYLIP software, and the CONSENSE programs determined the best-fitting tree. Multidimensional scaling (MDS) analyses were

Genetic Relations of Shia and Sunni Muslims / 431

Table 1.
Population Shia Muslims Sunni Muslims Jat Khatri Kurmi Thakur Brahmins Teli Yadav Pakistanis Afghans Georgians Jordanians Omanis Yemenis Iranians Iraqis

Populations Included in the Analyses

n 121 132 48 47 51 45 110 50 50 94 271 65 46 141 59 150 206 Country or Region North India North India North India North India North India North India North India North India North India Pakistan Afghanistan Georgia Jordan Oman Yemen Iran Iraq Description Islamic religious group Islamic religious group Hindu caste Hindu caste Hindu caste Hindu caste Hindu caste Hindu caste Hindu caste General population General population General population Arab population Arab population Arab population General population General population from central and southern provinces Saudi Arabian population General population Arabian population General population from the Van and Agri districts of eastern Anatolia Malaysian Malay population General population General population Hong Kong Chinese General population General population General population from Surabaya, eastern Java Swedish population General population General population, majority from Flanders region General population from Tuscany North and central Poland population Reference Eaaswarkhanth et al. (2009) Eaaswarkhanth et al. (2009) Tandon et al. (2002) Tandon et al. (2002) Tandon et al. (2002) Tandon et al. (2002) Dubey et al. (2009) Sarkar and Kashyap (2002) Ashma and Kashyap (2002) Shepard and Herrera (2006a) Berti et al. (2005) Shepard and Herrera (2006a) Shepard and Herrera (2006a) Shepard and Herrera (2006a) Shepard and Herrera (2006a) Shepard and Herrera (2006b) Barni et al. (2007)

Saudi Arabians Qataris Syrians Turks

94 133 121 116

Dubai Qatar Syria Turkey

Alshamali et al. (2005) Prez-Miranda et al. (2006) Abdin et al. (2003) Cakir et al. (2003)

Malaysians Taiwanese Japanese Chinese Thais Koreans Indonesians

210 597 526 325 210 231 105

Malaysia Taiwan Japan China Thailand Korea Indonesia

Seah et al. (2003) Wang et al. (2003) Hashiyadaa et al. (2003) Chan et al. (2005) Rerkamnuaychoke et al. (2006) Kim et al. (2003) Dobashi et al. (2005)

Swedes Greeks Belgians

296 298 222

Sweden Greece Belgium

Montelius et al. (2008) Snchez-Diz et al. (2008) Decorte et al. (2004)

Italians Poles

361 412

Italy Poland

Presciuttini et al. (2006) Czarny et al. (2005)

432 / eaaswarkhanth et al. executed using Neis DA genetic distances between pairs of populations and were plotted in a two-dimensional representation using SPSS software, version 12 (SPSS Inc. 2004). A locus-wise exact test of population differentiation was performed using Arlequin, version 3.1 (Excoffier et al. 2005), to analyze the extent of genetic diversity among the North Indian Muslims and their geographically closest neighboring Hindu populations (including several caste and tribal populations) at the analyzed 13 STR loci. Admix95 software, based on the gene identity method (Chakraborty 1985), was used to estimate the admixture proportions (m SE) of Shia and Sunni Muslim populations. For this, we chose three putative parental populations: (1) the geographically closest Indian Hindu population and a pool of populations from (2) Arabia and (3) Iran. The Arabian and Iranian parental populations are known to have historical significance pertaining to both Muslim populations.

Results
STR Diversity of Shia and Sunni Muslims. The allele-frequency distribution and statistical evaluation of the 15 autosomal STR loci analyzed from the Shia and Sunni Muslim populations are presented in Tables 2 and 3, respectively. In total, 142 alleles were observed in the Shia Muslim population with corresponding allele frequencies ranging from 0.004 to 0.455 (Table 2), whereas 147 alleles were observed in the Sunni Muslim population with corresponding allele frequencies ranging from 0.004 to 0.383 (Table 3). The TPOX and CSF1PO loci were found to have a maximum allele-frequency range with alleles 8 (0.455) and 12 (0.383) being the most frequent alleles in the Shia Muslim and Sunni Muslim populations, respectively. The power of discrimination ranged from a minimum of 0.8290 for TPOX to a maximum of 0.9640 for FGA in the Shia Muslim population and from 0.8730 for CSF1PO to 0.9710 for D18S51 in the Sunni Muslim population. The PIC values ranged from 0.6263 (TPOX) to 0.8952 (FGA) in Shia Muslims and from 0.6656 (CSF1PO) to 0.9054 (D18S51) in Sunni Muslims, indicating good informativeness of all STR markers. As expected, the most polymorphic loci were also the most discriminating: FGA (PD = 0.9640) in Shia Muslims and D18S51 (PD = 0.9710) in Sunni Muslims. The gene diversity values ranged from 0.6784 (TPOX) to 0.9027 (FGA) for Shia Muslims and from 0.7152 (CSF1PO) to 0.9120 (D18S51) for Sunni Muslims, with an average of 0.8022 and 0.8044, respectively. The combined power of exclusion and combined power of discrimination for all 15 STR loci were 0.9999 and greater than 0.99999, respectively, for both Shia and Sunni Muslim populations. The probability of matching value ( pM) and paternity index (PI) for all 15 STR loci of Shia Muslim and Sunni Muslim populations were 1 in 1.137 10-17 and 1 in 2.886 10-17 for pM and 7.502 10-4 and 2.559 10-5 for PI, respectively. The observed heterozygosity (Ho) values varied from 0.5833 (D18S51) to 0.8595 (VWA) in Shia Muslims and from 0.6818 (CSF1PO) to 0.8333 (D21S11) in Sunni Muslims with low expected heterozygosity (He) values in 11 out of 15 loci analyzed.

Genetic Relations of Shia and Sunni Muslims / 433 Among all the studied markers, D18S51 and FGA (p < 0.001) in both populations were found to depart significantly from Hardy-Weinberg expectations even after Bonferroni adjustment for number of loci tested (a = 0.05/15, or 0.0033). Furthermore, GenePop analysis showed considerable heterozygote deficiency for markers D18S51 and FGA in Shia Muslims and for marker D18S51 in Sunni Muslims. To determine the pattern of heterozygote excess and heterozygote deficiency, we analyzed data using Bottleneck software (Cornuet and Luikart 1996), which is based on the Wilcoxon test results of the two-phase mutation model with a 90% SMM (stepwise mutation model) proportion and 10% variance. Subsequently, the presence of undetected null alleles as a cause for heterozygote deficiency was examined with the help of Micro-Checker software (van Oosterhout et al. 2004) applying standard settings. We observed that the D18S51 and FGA loci in Shia Muslims and the D8S1179, D18S51, and D2S1338 loci in Sunni Muslims had signs of null alleles. Genetic Relationships and Admixture Estimates Among Populations: 13 STR Loci Analyses North Indian Muslim and Non-Muslim Populations. We performed an exact test of population differentiation between the Shia and Sunni Muslim populations and their corresponding geographically closest neighboring Hindu religious populations to test for genetic differences of statistical significance. The pairwise comparison test values are tabulated in Table 4. Shia Muslims showed significant differences from the Khatri, Kurmi, and Thakur at the D5S818 locus and from the Thakur at 10 out of 13 loci. A similar trend was also observed for Sunni Muslims, which significantly differed at 8 out of 13 loci from the Thakur and at the D5S818 and FGA loci from the Khatri, Kurmi, and Thakur. To determine the genetic relationships between the Shia and Sunni Muslims and their geographically closest neighboring Hindu religious populations, we generated an NJ tree based on Neis genetic distance (DA). In the NJ tree (Figure 2), Shia and Sunni Muslims separate from each other with a low incidence (bootstrap value = 16%). On the whole, with the geographically closest neighboring Hindu religious populations, both Muslim populations were positioned near the Yadav and Jat (bootstrap values = 26% and 35%, respectively). The MDS plot (Figure 3) shows the genetic distances (DA) between populations in two-dimensional space, and it essentially mirrors the phylogenetic analyses. Both Shia and Sunni Muslims dispersed closer in a downward diagonal position with the Yadav in the right-hand quadrant and with the Jat in the left-hand quadrant, whereas the Brahmin, Thakur, Khatri, Kurmi, and Teli were outliers. The stress value for the MDS plot is 0.047, and the proportion of variance accounted for by the corresponding distances of the scaled data is 99%. North Indian Muslims, Non-Muslims, and Populations of the Middle East, Europe, and East Asia. To ascertain the genetic affinities between North Indian Muslims along with their geographically closest neighboring Hindu religious

Table 2. Allele Frequencies and Statistical Parameters of Genetic and Forensic Interest Estimated for the 15 STR Loci in Shia Muslims (N = 121), Uttar Pradesh, India
TPOX D18S51 D3S1358 THO1 D13S317 D16S539 D2S1338 D5S818 FGA

Allele

D8S1179 D21S11 D7S820 CSF1PO D19S433 VWA

0.012 0.004 0.079 0.004 0.302 0.037 0.095 0.042 0.335 0.021 0.256 0.281 0.306 0.202 0.120 0.021 0.008 0.004 0.004 0.083 0.211 0.140 0.095 0.028 0.042 0.095 0.132

0.026 0.198 0.103 0.454 0.128 0.062 0.149

0.004 0.004 0.017

0.004 0.157 0.091

0.033 0.107 0.331 0.380 0.132 0.017

434 / eaaswarkhanth et al.

0.198

0.224

0.229

0.273 0.182 0.124 0.285 0.132 0.004

0.095 0.104

0.250 0.177

0.280 0.390

0.186 0.120 0.292

0.022

0.059

0.014 0.074 0.009 0.079 0.028 0.097

0.178

0.017

0.161

0.058 0.293 0.182 0.079 0.004 0.009 0.005 0.023 0.102 0.014

0.120 0.004 0.198 0.294 0.012 0.223 0.074 0.012 0.008

0.162 0.009 0.032

0.004

0.079 0.008 0.318 0.012 0.202 0.075 0.165 0.066 0.045 0.022 0.004

0.009 0.009 0.031 0.105 0.057 0.206 0.031

6 7 8 9 9.3 10 10.2 11 12 12.2 13 13.2 14 14.2 15 15.2 16 16.2 17 18 18.2 19 19.2 20 20.2 21 21.2

0.023 0.182 0.107 0.070 0.008 0.004

22 22.2 23 23.2 24 24.2 25 26 27 28 29 29.2 30 30.2 31 31.2 32.2 33.2 Ho He HWE PD PIC PE PI pM GD 0.793 0.815 0.365 0.936 0.789 0.587 2.420 0.064 0.812 0.860 0.810 0.456 0.927 0.780 0.714 3.560 0.073 0.807 0.694 0.681 0.007 0.829 0.626 0.419 1.640 0.171 0.678 0.583 0.857 0.000 0.948 0.860 0.271 1.200 0.052 0.871 0.686 0.748 0.056 0.893 0.702 0.407 1.590 0.107 0.744 0.769 0.782 0.747 0.911 0.743 0.542 2.160 0.089 0.778 0.752 0.807 0.143 0.927 0.777 0.513 2.020 0.073 0.804 PI, paternity index. pM, matching probability. GD, gene diversity. 0.843 0.811 0.751 0.930 0.781 0.681 3.180 0.070 0.807 0.818 0.870 0.005 0.959 0.852 0.633 2.750 0.041 0.866 0.744 0.719 0.442 0.857 0.668 0.499 1.950 0.143 0.716 0.632 0.898 0.000 0.964 0.895 0.331 1.360 0.036 0.903

0.050

0.096 0.061 0.123 0.066 0.092 0.009 0.070 0.031 0.004

0.802 0.849 0.615 0.955 0.826 0.602 2.520 0.045 0.845

0.025 0.132 0.190 0.004 0.136 0.038 0.054 0.149 0.202 0.070 0.835 0.858 0.256 0.955 0.838 0.665 3.030 0.045 0.855

0.759 0.808 0.006 0.924 0.794 0.525 2.070 0.076 0.819

0.712 0.716 0.926 0.867 0.681 0.447 1.740 0.133 0.727

Genetic Relations of Shia and Sunni Muslims / 435

Ho, observed heterozygosity. He, expected heterozygosity. HWE, Hardy-Weinberg p values.

PD, power of discrimination. PIC, polymorphic information content. PE, power of exclusion.

Table 3. Allele Frequencies and Statistical Parameters of Genetic and Forensic Interest Estimated for the 15 STR Loci in Sunni Muslims (N = 132), Uttar Pradesh, India
VWA 0.008 0.083 0.144 0.125 0.280 0.228 0.102 0.034 0.004 0.011 0.098 0.178 0.208 0.115 0.038 0.035 0.015 0.064 0.148 0.057 0.117 0.364 0.314 0.140 TPOX D18S51 D3S1358 THO1 D13S317 D16S539 D2S1338 D5S818 FGA

Allele

D8S1179 D21S11 D7S820 CSF1PO D19S433

0.030 0.004 0.004 0.087 0.004 0.080 0.008 0.121 0.087 0.212 0.245 0.205 0.095 0.027 0.031 0.092 0.023 0.019 0.046 0.193 0.155 0.008 0.108 0.326 0.100 0.265 0.053 0.027 0.004 0.061 0.345 0.034 0.238 0.325 0.076 0.023

0.011 0.197 0.053

0.008 0.008

0.008 0.356 0.166

0.227 0.178 0.174 0.254 0.167 0.004 0.201 0.068

436 / eaaswarkhanth et al.

0.201

0.261

0.186

0.072 0.086

0.277 0.178

0.306 0.382

0.178

0.023

0.098

0.163

0.008

0.008 0.050 0.031 0.115 0.015 0.112 0.008 0.157

0.163

0.004

0.087

6 7 8 9 9.3 10 10.2 11 12 12.2 13 13.2 14 14.2 15 15.2 16 16.2 17 18 19 20 20.2 21 21.2 22 22.2 23 0.303 0.023 0.234 0.057 0.121 0.053 0.094 0.023 0.004

0.008 0.008

0.045 0.068 0.015 0.189 0.019 0.148 0.008 0.220

0.004

0.091 0.034 0.011 0.004

0.004 0.004

23.2 24 25 25.2 26 26.2 27 28 29 29.2 30 30.2 31 31.2 32 32.2 33.2 34.2 35.2 Ho He HWE PD PIC PE PI pM GD 0.826 0.819 0.012 0.935 0.794 0.648 2.870 0.065 0.816 0.826 0.824 0.046 0.932 0.796 0.648 2.870 0.068 0.821 0.697 0.721 0.607 0.876 0.669 0.424 1.650 0.124 0.718 0.715 0.913 0.000 0.971 0.905 0.449 1.740 0.029 0.912 0.742 0.762 0.190 0.893 0.720 0.497 1.940 0.107 0.759 0.818 0.797 0.937 0.920 0.761 0.633 2.750 0.080 0.794 0.735 0.785 0.332 0.920 0.750 0.484 1.890 0.080 0.782 PI, paternity index. pM, matching probability. GD, gene diversity. 0.818 0.818 0.056 0.934 0.790 0.633 2.750 0.066 0.815 0.795 0.868 0.019 0.962 0.851 0.591 2.440 0.038 0.865 0.758 0.735 0.945 0.878 0.688 0.523 2.060 0.122 0.732 0.703 0.882 0.000 0.956 0.844 0.605 2.540 0.044 0.859

0.023 0.136 0.086 0.004 0.027 0.008 0.004

0.788 0.857 0.232 0.958 0.836 0.577 2.360 0.042 0.854

0.008 0.179 0.239 0.004 0.139 0.028 0.048 0.119 0.008 0.167 0.046 0.011 0.004 0.833 0.846 0.048 0.949 0.824 0.662 3.000 0.051 0.843

0.818 0.784 0.636 0.911 0.746 0.633 2.750 0.089 0.781

0.682 0.718 0.152 0.873 0.666 0.401 1.570 0.127 0.715

Genetic Relations of Shia and Sunni Muslims / 437

Ho, observed heterozygosity. He, expected heterozygosity. HWE, Hardy-Weinberg p values.

PD, power of discrimination. PIC, polymorphic information content. PE, power of exclusion.

Table 4. Population Differentiation Tests Between Shia and Sunni Muslims and Their Neighboring Hindu Religious Populations Based on 13 Autosomal STR Markers
VWA 0.19345 0.01083 0.69965 0.00000 0.18986 0.12421 0.45538 0.26479 0.01348 0.37751 0.00000 0.38511 0.15914 0.87546 0.74381 0.00000 0.78098 0.13696 0.37434 0.73673 0.39446 0.01626 0.46770 0.20700 0.21037 0.00051 0.04586 0.19561 0.37375 0.11240 0.46075 0.15041 0.01420 0.06584 0.51950 0.83632 0.16198 0.01214 0.00452 0.93497 0.09052 0.53162 0.66135 0.30175 0.24673 0.02128 0.00075 0.30334 0.48761 0.31941 0.00000 0.42350 0.02651 0.30423 0.24867 0.25904 0.14167 0.25777 0.10746 0.01259 0.00070 0.43592 0.17320 0.09734 0.11754 0.76314 0.08125 0.64707 0.54153 0.95868 0.95066 0.32418 0.01322 0.00433 0.74489 0.27746 0.40356 0.64806 0.17306 0.39702 0.03723 0.00013 0.26585 0.39858 0.84744 0.98280 0.52767 0.82955 0.67912 0.62416 0.83632 0.89968 0.97208 0.23742 0.73789 0.24638 0.56185 0.69281 TPOX D18S51 D3S1358 THO1 D13S317 D16S539 D5S818 0.00173 0.00000 0.00142 0.00000 0.12111 0.25141 0.38636 0.00054 0.00000 0.00041 0.00000 0.09869 0.54141 0.77679 FGA 0.23665 0.00343 0.08898 0.01994 0.06661 0.15707 0.15130 0.40052 0.00014 0.01975 0.00277 0.87455 0.24198 0.50069

438 / eaaswarkhanth et al.

Population Comparison 0.31012 0.18393 0.93698 0.06902 0.07743 0.60341 0.66239 0.18051 0.34389 0.91086 0.01598 0.11144 0.77736 0.84709

D8S1179

D21S11

D7S820

CSF1PO

Shia vs. Jat Khatri Kurmi Thakur Brahmin Teli Yadav Sunni vs. Jat Khatri Kurmi Thakur Brahmin Teli Yadav

0.70873 0.74821 0.07687 0.00797 0.59289 0.00009 0.43547

0.47394 0.15862 0.06649 0.00000 0.76342 0.52946 0.30286

0.12991 0.54241 0.63075 0.03619 0.91652 0.38192 0.64626

0.80016 0.73303 0.17240 0.07801 0.26642 0.00552 0.96134

0.87258 0.09024 0.12022 0.00014 0.99059 0.68632 0.88521

0.01418 0.25096 0.15664 0.02896 0.47863 0.10122 0.28563

Boldface values indicate a significant difference between the studied population and its neighboring population.

Genetic Relations of Shia and Sunni Muslims / 439

Figure 2. Neighbor-joining tree based on Neis DA genetic distances between Shia Muslims, Sunni Muslims, and non-Muslims generated from the 13 CODIS core STR markers.

populations and the pooled populations from the Middle East, Europe, and East Asia, we conducted phylogenetic analyses using the NJ and MDS methods based on Neis DA genetic distance. We observed a strong line of geographic demarcation in the worldwide consensus NJ tree (Figure 4). All the North Indian Muslims and non-Muslims clustered together with Pakistanis representing South Asia, whereas Middle Eastern, European, and East Asian populations grouped separately according to their geographic regions (Figure 4). In the MDS plot, Shia and Sunni Muslims along with North Indian nonMuslims and Pakistanis were disseminated in the upper quadrant of the plot, whereas the pooled populations from the Middle East, Europe, and East Asia were clustered in the lower quadrants of the two-dimensional plot (Figure 5). The stress value is 0.144, and the total variance accounted for is 92%. In the worldwide

440 / eaaswarkhanth et al.

Figure 3. Multidimensional-scaling plot depicting relationships between Shia Muslims, Sunni Muslims, and non-Muslims based on Neis DA genetic distances generated from the 13 CODIS core STR markers. The stress value is 0.047.

phylogenetic analyses, both the NJ tree and the MDS plot showed similar patterns of geographic separation. The genetic admixture estimates from the weighted least squares by gene identity method for the Shia and Sunni Muslim populations are shown in Table 5. Based on the allele frequencies, we calculated the admixture proportions with three putative parental populations: (1) the geographically closest Indian Hindu religious (caste) populations and a pool of populations from (2) Arabia and (3) Iran. Genetic contributions from the closest Hindu parental populations to Shia and Sunni Muslims were estimated at 50% and 56%, respectively, whereas the contributions from Iranians were 47% and 44%, respectively. We noticed low or no contribution from the Arabian gene pool. However, overall contributions from the parental populations to Shia and Sunni Muslims had positive correlation coefficients (R2) of 0.922 and 0.997, respectively (Table 5).

Discussion
Historical and anthropological facts suggest that Uttar Pradesh is the center for North Indian Muslims (Hollister 1988; Robb 2002; Schimmel 1982). Because the region was dominated by Muslim rulers during the medieval period, genetic

Genetic Relations of Shia and Sunni Muslims / 441

Figure 4. Neighbor-joining tree based on Neis DA genetic distances between Shia Muslims, Sunni Muslims, non-Muslims, and populations from the Middle East, East Asia, and Europe generated from the 13 CODIS core STR markers.

signatures of the invaders could be traced among the Muslim populations inhabiting the region (Papiha 1996). Therefore present-day North Indian Muslims may be the descendants of the Muslim invaders or the descendants of local converts. To substantiate the classical marker studies (Papiha 1996) and historical information (Hollister 1988; Robb 2002; Schimmel 1982), we assessed genetic diversity patterns, affinities, and level of admixture in Shia and Sunni Muslims of North India using a battery of commonly available 13 STR markers. The parameters of population genetics and forensic interest calculated on the basis of allele frequencies strongly support the use of the set of genetic markers used in the present study for personal identification testing. We observed that out of the 15 analyzed STR loci, 11 loci in both Muslim populations showed lower observed heterozygosity values than the expected heterozygosity values. This observation was predominantly noticeable for the D18S51 and FGA loci, which departed from Hardy-Weinberg expectations with Ho /He ratios of 0.68 and 0.70 for Shia and 0.78 and 0.79 for Sunni Muslims, respectively. These values are significantly distant from the theoretically expected values of 1 and are indicative of a relevant heterozygosity deficit.

442 / eaaswarkhanth et al.

Figure 5. Multidimensional-scaling plot portraying relationships between Shia Muslims, Sunni Muslims, non-Muslims, and populations from the Middle East, East Asia, and Europe. The stress value is 0.144. Right triangles, North Indian non-Muslim populations; diamonds, North Indian Muslim populations; hexagons, East Asian populations; isosceles triangles, Middle Eastern populations; squares, European populations. shm, Shia Muslims; sum, Sunni Muslims; jat, Jat; ktr, Khatri; kur, Kurmi; thr, Thakur; brh, Brahmins; tli, Teli; ydv, Yadav; grg, Georgians; jrd, Jordanians; omn, Omanis; yem, Yemenis; irn, Iranians; irq, Iraqis; sab, Saudi Arabians; qtr, Qataris; syr, Syrians; try, Turks; mal, Malaysians; tai, Taiwanese; jpn, Japanese; chn, Chinese; thi, Thais; kor, Koreans; ind, Indonesians; sdn, Swedes; grk, Greeks; bel, Belgians; ity, Italians; pld, Poles.

Consanguineous marriage is widely practiced among Muslims of this region. Such marriage preference varies by sect (Sunni 60.9% versus Shia 43.4%) and clan affiliation (Bittles and Hussain 2000). The scientific fact that homozygosity at genetic loci increases distinctly in populations practicing consanguinity seems to be reflected in the comparatively low observed heterozygosity values for most of the STR loci analyzed in both the Shia and Sunni Muslim populations. Thus the departures from Hardy-Weinberg expectations detected in the Shia and Sunni Muslim populations appear to be due to an excess of homozygotes over heterozygotes, which is likely the consequence of high consanguinity rates reported for these populations (Afzal 1984; Bittles and Hussain 2000). The regional phylogenetic assessment divulges the genetic proximity of Shia and Sunni Muslims to some of their geographically close Hindu religious

Genetic Relations of Shia and Sunni Muslims / 443

Table 5. Admixture Proportions ( Standard Error) Based on 13 Autosomal STR Markersa
Admixed Populations Parental Group Uttar Pradesh Iran Arabia R2 b Shia Muslims 0.496 0.082 0.473 0.080 0.031 0.085 0.922 Sunni Muslims 0.563 0.026 0.441 0.023 -0.004 0.018 0.997

a. The admixture proportion values are from Eaaswarkhanth et al. (2009). b. Correlation coefficient of admixture contributions from the parental groups.

populations, the Yadav and the Jat (see Figures 2 and 3). This is also evident from the population differentiation test, in which both Muslim populations did not vary significantly from the Yadav and the Teli at least for the markers analyzed (see Table 4). In the worldwide phylogeny, both Shia and Sunni Muslims were placed in the South Asian group (see Figure 4). We did not observe any genetic affinity with either the homelands of the Islamic invaders (i.e., Arabia and Iran) or any other regions of the world included in the analyses. However, a significant degree of genetic contribution from Iran to Shia (47%) and Sunni (44%) Muslims suggests genetic inputs from the region (see Table 5). Therefore our high-resolution analysis delineated the genetic affinities between Shia and Sunni Muslims from northern India and geographically targeted neighboring and global populations.
Acknowledgments We thank all the generous donors who participated in the present study and Faizan A. Khan for his help in sample collection. M. Eaaswarkhanth, B. Dubey, P. R. Meganathan, and S. Noor thank the Directorate of Forensic Science, Ministry of Home Affairs, Government of India, for their fellowships. Received 13 May 2009; revision accepted for publication 8 July 2009.

Literature Cited
Aarzoo, S. S., and M. Afzal. 2005. Gene diversity in some Muslim populations of North India. Hum. Biol. 77:343353. Abdin, L., I. Shimada, B. Brinkmann et al. 2003. Analysis of 15 short tandem repeats reveals significant differences between the Arabian populations from Morocco and Syria. Leg. Med. (Tokyo) 5:S150S155. Afzal, M. 1984. Effects of Consanguinity on Reproductive Fitness and Certain Behavioral Traits Among Bihar Muslims. Ph.D. thesis. Bhagalpur University, Bihar. Alshamali, F., A. Q. Alkhayat, B. Budowle et al. 2005. STR population diversity in nine ethnic populations living in Dubai. Forensic Sci. Int. 152:267279. Applied Biosystems. 2001. Applied Biosystems AmpFlSTR Identifiler PCR Amplification Kit Users Manual: Instruction for Use of Products. Foster City, CA: Applied Biosystems.

444 / eaaswarkhanth et al.

Ara, G., Y. H. Siddique, T. Beg et al. 2008. Gene diversity among some Muslim populations of western Uttar Pradesh, India. Anthropologist 10:59. Ashma, R., and V. K. Kashyap. 2002. Genetic polymorphism at 15 STR loci among three important subpopulations of Bihar, India. Forensic Sci. Int. 130:5862. Barni, F., A. Berti, A. Pianese et al. 2007. Allele frequencies of 15 autosomal STR loci in the Iraq population with comparisons to other populations from the Middle Eastern region. Forensic Sci. Int. 167:8792. Berti, A., F. Barni, A. Virgili et al. 2005. Autosomal STR frequencies in Afghanistan population. J. Forensic Sci. 50:14941496. Bittles, A. H., and R. Hussain. 2000. An analysis of consanguineous marriage in the Muslim population of India at regional and the state levels. Ann. Hum. Biol. 27:163171. Cakir, A. H., A. Celebioglu, S. Altunbas et al. 2003. Allele frequencies for 15 STR loci in Van-Agri districts of the eastern Anatolia region of Turkey. Forensic Sci. Int. 135:6063. Census of India. 2001. Census 2001. www.censusindia.net/. Chakraborty, R. 1985. Gene identity in racial hybrids and estimation of admixture rates. In Genetic Microdifferentiation in Man and Other Animals, J. V. Neel and Y. R. Ahuja, eds. Delhi: Delhi University and Indian Anthropological Association, 171180. Chan, K. M., C. T. Chiu, P. Tsui et al. 2005. Population data for the Identifiler 15 STR loci in Hong Kong Chinese. Forensic Sci. Int. 152:307309. Cornuet, J. M., and G. Luikart. 1996. Description and power analysis of two tests for detecting recent population bottlenecks from allele frequency data. Genetics 144:20012014. Czarny, J., T. Grzybowski, M. V. Derenko et al. 2005. Genetic variation of 15 STR loci (D3S1358, VWA, FGA, TH01, TPOX, CSF1PO, D5S818, D13S317, D7S820, D16S539, D2S1338, D8S1179, D21S11, D18S51, and D19S433) in populations of north and central Poland. Forensic Sci. Intl. 147:97100. Decorte, R., M. Engelen, L. Larno et al. 2004. Belgian population data for 15 STR loci (AmpFlSTR SGM Plus and AmpFlSTR Profiler PCR amplification kit). Forensic Sci. Int. 139:211213. Dobashi, Y., A. Kido, N. Fujitani et al. 2005. STR data for the AmpFlSTR Identifiler loci in Bangladeshi and Indonesian populations. Leg. Med. (Tokyo) 7:222226. Dubey, B., P. R. Meganathan, M. Eaaswarkhanth et al. 2009. Forensic STR profile of two endogamous populations of Madhya Pradesh, India. Leg. Med. (Tokyo) 11:4144. Eaaswarkhanth, M., B. Dubey, P. R. Meganathan et al. 2009. Diverse genetic origin of Indian Muslims: Evidence from autosomal STR loci. J. Hum. Genet. 54(6):340348. Eaaswarkhanth, M., T. S. Vasulu, and I. Haque. 2008. Genetic affinity between diverse ethno-religious communities of Tamil Nadu, India: A microsatellite study. Hum. Biol. 80:601609. Excoffier, L., G. Laval, and S. Schneider. 2005. Arlequin ver. 3.0: An integrated software package for population genetics data analysis. Evol. Bioinform. 1:4750. http://cmpg.unibe.ch/software/ arlequin3. Felsenstein, J. 1993. Phylogeny Inference Package (PHYLIP), version 3.6c/. Distributed by the author, Department of Genetics, University of Washington, Seattle. Hashiyadaa, M., Y. Itakurab, T. Nagashima et al. 2003. Polymorphism of 17 STRs by multiplex analysis in Japanese population. Forensic Sci. Int. 133:250253. Hollister, J. N. 1988. Islam and Shias Faith in India. New Delhi: Kanishka. Kim, Y.-L., J.-H. Hwang, Y.-J. Kim et al. 2003. Allele frequencies of 15 STR loci using AmpFlSTR Identifiler kit in a Korean population. Forensic Sci. Int. 136:9295. Lal, B. B. 2005. The Homeland of the Aryans: Evidence of Rigvedic Flora and Fauna and Archaeology. New Delhi: Aryan Books International. Lanchbury, J. S., S. S. Agarwal, and S. S. Papiha. 1996. Genetic differentiation and population structure of some occupational caste groups in Uttar Pradesh, India. Hum. Biol. 68:655678. Lui, K., and S. V. Muse. 2005. Powermarker: Integrated analysis environment for genetic marker data. Bioinformatics 21:21282129. http://statgen.ncsu.edu/powermarker/.

Genetic Relations of Shia and Sunni Muslims / 445

Montelius, K., A. O. Karlsson, and G. Holmlund. 2008. STR data for the AmpFlSTR Identifiler loci from Swedish population in comparison to European, as well as with non-European population. Forensic Sci. Int. Genet. 2:e49e52. Papiha, S. S. 1996. Genetic variation in India. Hum. Biol. 68:607628. Prez-Miranda, A. M., M. A. Alfonso-Snchez, J. A. Pea et al. 2006. Qatari DNA variation at a crossroad of human migrations. Hum. Hered. 61:6779. Presciuttini, S., N. Cerri, S. Turrina et al. 2006. Validation of a large Italian database of 15 STR loci. Forensic Sci. Int. 156:266268. Raymond, M., and F. Rousset. 1995. GenePop (version 1.2): Population genetics software for exact tests and ecumenicism. J. Hered. 86:248249. http://genepop.curtin.edu.au. Rerkamnuaychoke, B., T. Rinthachai, J. Shotivaranon et al. 2006. Thai population data on 15 tetrameric STR loci: D8S1179, D21S11, D7S820, CSF1PO, D3S1358, TH01, D13S317, D16S539, D2S1338, D19S433, VWA, TPOX, D18S51, D5S818, and FGA. Forensic Sci. Intl. 158:234237. Robb, P. 2002. A History of India. Hampshire, U.K.: Palgrave Macmillan. Sambrook, J., E. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press. Snchez-Diz, P., P. G. Menounos, A. Carracedo et al. 2008. 16 STR data of a Greek population. Forensic Sci. Int. Genet. 2:e71e72. Sarkar, N., and V. K. Kashyap. 2002. Genetic diversity at two pentanucleotide STR and thirteen tetranucleotide STR loci by multiplex PCR in four predominant population groups of central India. Forensic Sci. Int. 128:196201. Schimmel, A. 1982. Islam in India and Pakistan. Leiden, Netherlands: Brill Academic. Seah, L. H., N. H. Jeevan, M. I. Othman et al. 2003. STR data for the AmpFlSTR Identifiler loci in three ethnic groups (Malay, Chinese, Indian) of the Malaysian population. Forensic Sci. Int. 138:134137. Shepard, E. M., and R. J. Herrera. 2006a. Genetic encapsulation among Near Eastern populations. J. Hum. Genet. 51:467476. Shepard, E. M., and R. J. Herrera. 2006b. Iranian STR variation at the fringes of biogeographical demarcation. Forensic Sci. Int. 158:140148. SPSS Inc. 2004. Statistical Package for the Social Sciences (SPSS) Software 12.0.1: Evaluation Version for Windows. Chicago: SPSS Inc. Tandon, M., R. Trivedi, and V. K. Kashyap. 2002. Genomic diversity at 15 fluorescent labeled short tandem repeat loci in few important populations of state of Uttar Pradesh, India. Forensic Sci. Intl. 128:190195. van Oosterhout, C., W. F. Hutchinson, D. P. M. Wills et al. 2004. Micro-Checker: Software for identifying and correcting genotyping errors in microsatellite data. Mol. Ecol. Notes 4:535538. www.microchecker.hull.ac.uk/. Wang, C. W., D. P. Chen, C. Y. Chen et al. 2003. STR data for the AmpFlSTR SGM Plus and Profiler loci from Taiwan. Forensic Sci. Int. 138:119122.