Copyright  2002 by the Genetics Society of AmericaLarger Genetic Differences Within Africans Than BetweenAfricans and EurasiansNing Yu,* Feng-Chi Chen,*,†Satoshi Ota,* Lynn B. Jorde,‡Pekka Pamilo,§Laszlo Patthy,**Michele Ramsay,††Trefor Jenkins,** Song-Kun Shyue‡‡and Wen-Hsiung Li*,1*Department of Ecology and Evolution, University of Chicago, Chicago, Illinois 60637,†Department of Life Science, National Tsing HuaUniversity, Hsinchu, 300 Taiwan,‡Department of Human Genetics, University of Utah, Salt Lake City, Utah 84112,§Department of Biology,University of Oulu, 90014 Oulu, Finland, **Institute of Enzymology, Biological Research Center, Hungarian Academy ofSciences, H-1518 Budapest, Hungary,††Department of Human Genetics, South African Institute for MedicalResearch and University of the Witwatersrand, Johannesburg, 2050 South Africa and‡‡Instituteof Biomedical Sciences, Academia Sinica, Taipei, 115 TaiwanManuscript received August 30, 2001Accepted for publication February 19, 2002ABSTRACTThe worldwide pattern of single nucleotide polymorphism (SNP) variation is of great interest to humangeneticists, population geneticists, and evolutionists, but remains incompletely understood. We studiedthe pattern in noncoding regions, because they are less affected by natural selection than are coding regions.Thus, it can reflect better the history of human evolution and can serve as a baseline for understanding themaintenance of SNPs in human populations. We sequenced 50 noncoding DNA segments each ⵑ500 bplong in 10 Africans, 10 Europeans, and 10 Asians. An analysis of the data suggests that the samplingscheme is adequate for our purpose. The average nucleotide diversity (␲) for the 50 segments is only0.061% ⫾ 0.010% among Asians and 0.064% ⫾ 0.011% among Europeans but almost twice as high(0.115% ⫾ 0.016%) among Africans. The African diversity estimate is even higher than that betweenAfricans and Eurasians (0.096% ⫾ 0.012%). From available data for noncoding autosomal regions (totallength ⫽ 47,038 bp) and X-linked regions (47,421 bp), we estimated the ␲-values for autosomal regionsto be 0.105, 0.070, 0.069, and 0.097% for Africans, Asians, Europeans, and between Africans and Eurasians,and the corresponding values for X-linked regions to be 0.088, 0.042, 0.053, and 0.082%. Thus, Africansdiffer from one another slightly more than from Eurasians, and the genetic diversity in Eurasians is largelya subset of that in Africans, supporting the out of Africa model of human evolution. Clearly, one mustspecify the geographic origins of the individuals sampled when studying ␲ or SNP density.THERE has been much interest in single nucleotide serve as a baseline for understanding the maintenanceof SNPs in human populations. For these reasons, wepolymorphisms (SNPs) in human populations be-cause such data are useful for studying human evolution have obtained SNP data from 50 noncoding regions inAfricans, Europeans, and Asians and have estimatedand the mechanism of maintenance of genetic variabil-ity in human populations and for identifying genes asso- the levels of nucleotide diversity within and betweenpopulations. To strengthen our conclusions, we haveciated with complex disease (Harding et al. 1997; Nick-erson et al. 1998; Zietkiewicz et al. 1998; Chargill et also used data from GenBank and the literature.al. 1999; Halushka et al. 1999; Harris and Hey 1999;Jaruzelska et al. 1999; Kaessmann et al. 1999; RiederMATERIALS AND METHODSet al. 1999; Nachman and Crowell 2000). The muchinterest notwithstanding, the worldwide pattern of SNPDNA samples: The 10 Africans used were 1 Biaka Pygmy,1 Mbuti Pygmy, 1 Ghanaian, 1 Kikuyu, 1 !Kung, 1 Luo, 2variation remains incompletely understood. This is espe-Nigerians (Yuroba and Rivers), 1 South African Bantu speaker,cially true for noncoding regions because such regionsand 1 Zulu (also a South African Bantu speaker); the 10are medically less interesting. However, data from non-Europeans were 1 Finnish, 1 French, 1 German, 1 Hungarian,coding regions are less affected by natural selection1 Italian, 1 Portuguese, 1 Russian, 1 Spanish, 1 Swedish, andthan data from coding regions and so can reflect more1 Ukranian; and the 10 Asians were 1 Cambodian, 2 Chinese(North and South), 1 Han Taiwanese, 2 Indians (Punjab andaccurately the history of human evolution. Moreover,Bengal), 1 Japanese, 1 Mongolian, 1 Vietnamese, and 1 Yakut.the pattern of SNP variation in noncoding regions canAs every segment studied is autosomal, the number of se-quences studied for each segment is 60 (20 for each continentstudied).Selection of DNA segments: Fifty noncoding, nonrepetitive1Corresponding author: Department of Ecology and Evolution, Univer-genomic segments (each ⵑ1 kb), which covered almost allsity of Chicago, 1101 E. 57th St., Chicago, IL 60637.E-mail: [email protected] autosomes, were selected randomly with reference to the Ge-Genetics 161: 269–274 (May 2002)270 N. Yu et al.nome Channel (http://genome.ornl.gov/GCat/species.shtml);two randomly chosen sequences in a population. Thesee Chen and Li (2001) for details. All of them were chosen␲-value fluctuates greatly among the segments studiedto avoid coding or close linkage to any coding regions. In(Table 1). The range of ␲ is from 0 (5 segments) to 0.27%each segment and its nearby regions there was no registeredin the total sample, from 0 (5 segments) to 0.58% ingene in GenBank and no potential coding region was detectedby either GenScan or GRAIL-EXP.the African sample, from 0 (19 segments) to 0.27% inPCR amplification and DNA sequencing: Touchdown PCRthe Asian sample, and from 0 (18 segments) to 0.29%(Don et al. 1991) was used and the reactions were carried outin the European sample. Such large fluctuations arefollowing the conditions described (Zhao et al. 2000). Thenot surprising because the nucleotide diversity in a shortPCR products were purified by the Wizard PCR Preps DNADNA region is subject to strong stochastic effects. Inpurification resin kit (Promega, Madison, WI). Sequencingreaction was performed according to the protocol of ABI Prismaddition, variation in ␲ may also arise from differentBigDye Terminator sequencing kits (Perkin-Elmer, Norwalk,mutation rates among different segments, although weCT) modified by quarter reaction. The extension productsfound no correlation between ␲ and the divergence be-were purified by Sephadex G-50 (DNA grade; Pharmacia, Pis-tween human and