A microsatellite-based analysis for the detection of selection on BTA1 and BTA20 in northern Eurasian cattle (Bos taurus) populations

RESEARCH Open Access A mic osa elli e-based analysis o he de ec ion o selec ion on BTA1 and BTA20 in no he n Eu asian ca le (Bos au us) popula ions Meng-Hua Li, Te hi Iso-Tou u, Hannele Lau én, Juha Kan anen * Abs ac Backg ound: Mic osa elli es su ounding unc ionally impo an candida e genes o quan i a i e ai loci ha e ecei ed a en ion as p oxy measu es o polymo phism le el a he candida e loci hemsel es. In ca le, selec ion o economically impo an ai s is a long- e m s a egy and i has been epo ed ha mic osa elli es a e linked o hese impo an loci. Me hods: We ha e in es iga ed he a ia ion o se en mic osa elli es on BTA1 (Bos au us au osome 1) and 16 on BTA20, using bo ine popula ions o ypical p oduc ion ypes and ho n s a us in no he n Eu asia. Gene ic a iabili y o hese loci and linkage disequilib ium among hese loci we e compa ed wi h hose o 28 mic osa elli es on o he bo ine ch omosomes. Fou di e en es s we e applied o de ec molecula signa u es o selec ion. Resul s: No ma ked di e ence in locus a iabili y was ound be ween mic osa elli es on BTA1, BTA20 and he o he ch omosomes in e ms o di e en di e si y indices. A e age D′ alues o pai wise syn enic ma ke s (0.32 and 0.28 ac oss BTA 1 and BTA20 espec i ely) we e signi ican ly (P< 0.05) highe han o non-syn enic ma ke s (0.15). The Ewens-Wa e son es , he Beaumon and Nichol’s modi ied equen is es and he Bayesian F ST - es indica ed ele a ed o dec eased gene ic di e en ia ion, a SOD1 and AGLA17 ma ke s espec i ely, de ia ing signi ican ly (P< 0.05) om neu al expec a ions. Fu he mo e, lnRV, lnRH and lnRθ’s a is ics we e used o he pai wise popula ion compa ison es s and we e signi ican ly less a iable in one popula ion ela i e o he o he , p o iding addi ional e idence o selec ion signa u es o wo o he 51 loci. Mo eo e , he h ee Finnish na i e popula ions showed e idence o subpopula ion di e gence a SOD1 and AGLA17. Ou da a also indica e signi ican in e genic linkage disequilib ium a ound he candida e loci and sugges ha hi chhiking selec ion has played a ole in shaping he pa e n o obse ed linkage disequilib ium. Conclusion: Hi chhiking due o igh linkage wi h alleles a candida e genes, e.g. he POLL gene, is a possible explana ion o his pa e n. The po en ial impac o selec i e b eeding by man on ca le popula ions is discussed in he con ex o selec ion e ec s. Ou esul s also sugges ha a p ac ical app oach o de ec loci unde selec ion is o simul aneously apply mul iple neu ali y es s based on di e en assump ions and es ima ions. Backg ound Expec a ion o neu ali y ega ding he mu a ion-d i equilib ium o mic osa elli e a ia ion is no always alid due o demog aphic changes, including gene ic bo lenecks and admix u e (e.g. [1,2]), and selec ion a linked si es (e.g. [3,4]). In con as o demog aphic p o- cesses, which a ec he en i e genome, selec ion ope a es a speci ic si es associa ed wi h pheno ypic ai s, such as impo an quan i a i e ai loci (QTLs) and candida e genes. Selec ion lea es i s signa u e in he ch omosomal egions su ounding he si es, whe e sig- ni ican ly educed o ele a ed le els o gene ic a ia ion can be main ained a linked neu al loci. Thus, selec ion no only a ec s he selec ed si es bu also linked neu al loci and he oo p in s o selec ion ac ing on speci ic unc ional loci can be de ec ed by geno yping poly- mo phic mic osa elli es in he adjacen non-coding egions [5]. * Co espondence: [email p o ec ed] Bio echnology and Food Resea ch, MTT Ag i ood Resea ch Finland, FI-31600 Jokioinen, Finland Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Gene ics Selec ion E olu ion © 2010 Li e al; licensee BioMed Cen al L d. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/2.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Di e en s a is ical me hods ha e been de eloped o iden i y ou lie loci unde he in luence o selec ion [6-13] and adap a ions ha e been a emp ed o imp o e he o iginal me hods o Lewon in and K akaue [14], which ha e been c i icized because o hei sensi i i y o popula ion s uc u e and his o y (e.g. [15]). Ne e heless, ecen s udies ha e shown somewha inconsis en esul s ob ained by applying he abo e s a is ical es s o he same da a (e.g. [7,12,16,17]). The Lewon in- K akaue es [14] is he oldes o hese mul ilocus-compa ison me hods. B oadly speaking, hese me hods a e de i ed by using one o he wo gene al app oaches de ailed below. The i s app oach is o de elop me hods wi h Lewon in and K akaue s’o iginal idea and o use he dis ibu ion o es ima es o gene ic di e en ia ion coe i- cien F ST and di e si y pa ame e s om indi idual gene ic loci o de ec he e ec s o selec ion, he ea e e med he F ST -based app oach, such as he FDIST p o- g am-based me hod [9], Bayesian eg ession [12], and popula ion-speci ic [7] me hods. Schlö e e and collea- gues ha e p oposed al e na i e mul ilocus simula ion- based es s ha use summa y s a is ics o he han F ST , such as he ln RV [10], he ln RH [6], and he ln Rθ’ [13] es s. These es s in ol e conside ing he idea o a ‘selec i e sweep’ ha a ises om na u al and a i icial selec ion, and ecen gene ic exchanges d i en by he selec i e sweep lea e a eco d o “gene ic signa u e”in he genome co e ing he selec ed si es and hei linked neu al loci. Gi en ha mic osa elli e loci associa ed wi h a ecen selec i e sweep di e om he emainde o he genome, hey a e expec ed o all ou side he dis- ibu ion o neu al es ima es o ln RV, ln RH o ln Rθ’ alues. As e iewed by [18-20], all he me hods ha e po en ial ad an ages and d awbacks, which can be due o di e en unde lying assump ions ega ding he demo- g aphic and mu a ional models on which hey a e based, as well as on unce ain y associa ed wi h he obus ness o he app oaches. The ecen inc eased a ailabili y o la ge genomic da a se s and he iden i ica ion o a ew genes o loci as he a ge s o domes ica ion o subsequen gene ic imp o e- men in ca le ha e enewed he in es iga ion o he genomic e ec s o selec ion. Candida e genes and QTL ha e been desc ibed on bo h BTA1 [21-25] and BTA 20 [26]. On BTA1, he POLL gene, cha ac e ized by wo alleles: P(polled) dominan o e H(ho n), is esponsible o he polled (i.e. ho nless) and ho n pheno ypes in ca - le and has been subjec ed o bo h na u al and a i icial selec ion. Geo ges e al. [21] ha e demons a ed gene ic linkage be ween he POLL gene and wo mic osa elli es, GMPOLL-1 and GMPOLL-2. These loci a e syn enic o he highly conse ed gene o supe oxide dismu ase 1 (SOD1). In addi ion, in a ious b eeds he POLL gene has been ound o be linked o he mic osa elli es TGLA49,AGLA17,INRA212 and KAP8, loca ed in he cen ome ic egion o BTA1 close o he SOD1 locus [22,23,25]. To da e, on BTA20 se e al QTL and candi- da e genes ha e been epo ed e.g. g ow h ho mone and p olac in ecep o genes [27] a ec ing con o ma ion and milk p oduc ion ai s, such as body dep h (e.g. [28]), udde (e.g. [29]), udde a achmen (e.g. [30]), milk yield (e.g. [31]), a pe cen age (e.g. [28]), and especially p o- ein con en (e.g. [28-30]). In his s udy on Bos au us, we p esen mic osa elli e da a using a ela i ely la ge numbe o loci han p e- iously epo ed, which mainly included he 30 mic osa- elli e ma ke s ecommended by he In e na ional Socie y o Animal Gene ics (ISAG)/Food and Ag icul- u e O ganiza ion o he Uni ed Na ions (FAO) wo king g oup (e.g. [2,24]; bu see also [32]). Among he 51 mic osa elli es geno yped on 10 ep esen a i e ca le popula ions o di e en o igins (na i e and mode n comme cial) and ho n s a uses (polled and ho ned) in he no he n e i o y o he Eu asian subcon inen , se en we e on BTA1 and 16 on BTA20. We applied ou es s o de ec molecula signa u es o selec ion, anging om es s o loci ac oss popula ions and he ecen ly p oposed pai wise popula ion es s using a dynamically adjus ed numbe o linked mic osa elli es [13]. We compa ed he consis ency o he di e en neu- ali y es s a ailable o iden i y loci unde selec ion in he no h Eu asian ca le popula ions in es iga ed he e. Ma e ials and me hods Popula ion samples and gene ic ma ke s Mic osa elli e da a om 10 di e en ca le (Bos au us) popula ions including 366 indi iduals we e analyzed. Finnish popula ions we e ep esen ed by Finnish Ay shi e (mode n comme cial, ho ned, n=40),Finnish Hols ein-F iesian (mode n comme cial, ho ned, n=40), Eas e n Finnca le (na i e, mos ly polled, n=31), Wes e n Finnca le (na i e, mos ly polled, n=37),and No he n Finnca le (na i e, mos ly polled, n= 26). We we e able o in e ence he he e ozygo ic s a us a he POLL locus in 19 pheno ypically polled ca le o he h ee Finnish na i e popula ions, on he basis o hei o sp ing/pa en pheno ypes. In addi ion, he e we e 19 animals ho ned ( ecessi e homozygo ic) in he Finnish na i e popula ions. Is oben (na i e, ho ned, n= 40), Yaku ian (na i e, ho ned, n= 51), and Kholmogo y (na i e, ho ned, n= 32) ca le we e sampled in Russia. Uk ainian G ey (na i e, ho ned, n=30)andDanish Je sey (mode n comme cial, ho ned, n= 39) we e sampled in Uk aine and Denma k, espec i ely. Du ing sample collec ion, he pedig ee in o ma ion and he he dsman’s knowledge we e used o ensu e he animals we e un ela ed. Addi ional in o ma ion on hese popula- ions has been epo ed in p e ious publica ions [2,33]. Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 2 o 14 Geno ypes o he 51 mic osa elli es we e used ( o de ails on he mic osa elli es, see [33-35]) among which da a o he 30 ma ke s om he panel o loci ecom- mended o gene ic di e si y s udies in ca le h p:// www.p ojec s. oslin.ac.uk/cdi /ma ke s.h ml we e aken om he li e a u e [2]. The 23 mic osa elli es (21 new ones and wo om he ecommended panel) on BTA1 and BTA20 we e chosen on he basis o hei icini y o genes and QTL, which could be conside ed as candida e loci o selec ion because o hei assumed in ol emen in he polled/ho ned pheno ype [22] and in milk yield and body composi ion [35]. De ails o he p ime s and mic osa elli e analysis p o ocols can be ound in CaD- Base h p://www.p ojec s. oslin.ac.uk/cdi /ma ke s.h ml and[34].In hiss udy,GHRJA.UP,5′- GGTTCGTTATGGAGGCAATG-3′,andGHRJA.DN, 5′-GTCACCGCTGGCAGTAGAT-3′p ime s we e designed based on he sequence o he p omo e egion o he g ow h ho mone ecep o gene [35] con aining mic osa elli e GHRJA. Danish Je sey animals we e ana- lyzed only a 41 loci (see Table 1). A ull lis o he loci s udied and hei ch omosomal and genomic loca ions, as well as popula ion and basic s a is ics, a e a ailable in Table 1. Mic osa elli e a iabili y measu es and es o linkage disequilib ium Mic osa elli e a iabili y, expec ed he e ozygosi y (H EXP ), allelic ichness (A R ), and Wei and Cocke ham’sF ST [36], we e es ima ed wi h he FSTAT p og am, e sion 2.9.3.2 [37]. The D′me ic used o es ima e he LD was calcula ed using Mul iallelic In e allelic Disequilib ium Analysis So wa e (MIDAS; [38]). Values o D′we e calcula ed o all syn enic ma ke pai s on BTA1 and BTA20 ac oss he popula ions. A mo e de ailed desc ip ion o he es ima ion o D′can be ound in [39]. The s a is ical signi icance o he obse ed associa ion be ween pai s o alleles unde he null hypo hesis o andom allelic asso men was es ed using a Mon e-Ca lo app oxima- ion o Fishe ’s exac es as implemen ed in he so - wa e ARLEQUIN [40] using a Ma ko chain ex ension o Fishe ’sexac es o R×Ccon ingency ables [41]. A o al o 100 000 al e na i e ables we e explo ed wi h he Ma ko chain and p obabili ies we e ypically es i- ma ed wi h a s anda d e o o < 0.001. Es ima ion o he D′me ic o LD and es s o hei signi icance we e conduc ed only in h ee Finnish na i e b eeds, i.e. No he n Finnca le, Eas e n Finnca le and Wes e n Finnca le. The g aphic summa y o he signi icance o LD de e mina ions was displayed using he HaploView p og am, e sion 4.0 [42]. Fishe ’s exac es s in he GENEPOP 4.0 [43] we e applied o assess LD de e mi- na ions be ween all locus pai s ac oss he sample. Tes s o de ec loci unde selec ion ac oss popula ions Possible depa u es om he s anda d neu al model o molecula e olu ion - po en ially e ealing demog aphic e en s o he exis ence o selec i e e ec s a ce ain loci - we e examined o each locus using he Ewens- Wa e son es [44,45] and he Beaumon and Nichols’s modi ied equen is me hod [9], as well as a mo e obus Bayesian es [12]. The Ewens-Wa e son es o neu ali y was pe - o med wi h he ARLEQUIN p og am [40] assuming an in ini e allele mu a ion model. To ob ain su icien p ecision wi h his es , he p obabili y was eco ded as hemeano 20independen epea so 1,000simula- ions. The equen is me hod used was ha p oposed by [9], u he de eloped by [12], and implemen ed in he FDIST2 p og am h p://www. ubic. dg.ac.uk/~mab/ so wa e.h ml, a cu en ly dis ibu ed e sion o he o iginal FDIST p og am as desc ibed by [12]. FDIST2 calcula es θ, Wei & Cocke ham’s [36] es ima o o di e si y o each locus in he sample. Coalescen simula ions a e hen pe o med o gene a e da a se s wi h a dis ibu ion o θcen e ed on he empi ical es i- ma es. Then, he quan iles o he simula ed F ST wi hin which he obse ed F ST ’s elland heP- alues o each locus we e de e mined. Ini ially an island model o popula ion di e en ia ion was used and he p ocedu e epea ed 50,000 imes o gene a e 95% con idence in e als o neu al di e en ia ion and o es ima e P- alues o depa u e o he loci om hese expec a- ions. Simula ion pa ame e s we e unde an in ini e allele mu a ion model o 100 demes, 10 sample popu- la ions, sample sizes o 100, and a weigh ed F ST simila o he immed mean F ST calcula ed om he empi i- cal dis ibu ion. Compu ed by emo ing he 30% high- es and lowes F ST alues obse ed in he empi ical da a se , he immed mean F ST is an es ima e o he a e age “neu al”F ST alue unin luenced by ou lie loci (see [46]). This me hod p o ides e idence o selec ion by looking o ou lie s wi h highe /lowe obse ed F ST - alues, con olling o P- alues [12]. The app oach is ai ly obus ega ding a ia ion in mu a- ion a e be ween loci, sample size, and whe he popu- la ions a e a equilib ium o no [9]. Beaumon & Balding’s [12] hie a chical-Bayesian me hod was pe o med using he BAYESFST p og am h p://www. eading.ac.uk/S a is ics/gene ics/so wa e. h ml package, which gene a es 2,000 Ma ko chain Mon e Ca lo (MCMC) simula ed loci on he basis o he dis ibu ion o F ST gi en he da a. The me hod combines in o ma ion o e loci and popula ions in o de o simul aneously es ima e F ST a he i h locus and he j h popula ion, F ST (i,j), o all iloci and j popula ions. A hie a chical model is implemen ed o F ST (i,j)as Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 3 o 14 Table 1 Summa y o he mic osa elli es and basic popula ion gene ic es ima es o he mic osa elli es Locus BTA Genomic posi ion (bp) A R H E F IS FDIST2 es Ewens-Wa e son es s a s ends F ST PF OBS F EXP P H P E AGLA17 1 641402 641615 1.37 0.08 -0.049 0.017 0.010** 0.907 0.754 0.978* 0.976* DIK4591 1 1704734 1705228 2.60 0.32 0.064 0.128 0.660 0.467 0.442 0.844 0.622 DIK1044 1 2829429 2829737 4.86 0.70 0.015 0.118 0.631 0.324 0.329 0.136 0.243 SOD1 1 2914373 2915349 4.78 0.65 0.083 0.173 0.968* 0.331 0.379 0.037* 0.047* DIK5019 1 3900549 3900808 5.42 0.59 0.190 0.164 0.954* 0.381 0.380 0.005** 0.008** BMS2321 1 10949260 10949302 3.58 0.45 0.154 0.094 0.410 0.429 0.486 0.424 0.052 BM1824 1 122531990 122532171 3.95 0.72 -0.083 0.122 0.655 0.450 0.487 0.030* 0.231 TGLA304 20 11460907 11460992 3.30 0.49 0.113 0.114 0.573 0.497 0.531 0.237 0.238 BMS1754 20 18439757 18439877 3.47 0.58 0.014 0.094 0.384 0.503 0.536 0.153 0.126 NRDIKM033 20 15598470 15598176 5.20 0.75 -0.004 0.098 0.372 0.234 0.213 0.415 0.466 ILSTS068 20 21675187 21675451 2.07 0.25 0.095 0.146 0.760 0.734 0.751 0.383 0.223 TGLA126 20 21808628 21808745 6.27 0.71 -0.009 0.079 0.170 0.493 0.443 0.085 0.057 BMS2461 20 25278607 25278662 4.83 0.62 0.028 0.180 0.985* 0.227 0.246 0.453 0.760 BMS1128 20 26364064 26364112 3.54 0.52 0.032 0.109 0.534 0.472 0.446 0.503 0.203 BM713 20 26977228 26977280 3.36 0.62 -0.074 0.162 0.907 0.439 0.486 0.197 0.674 DIK2695 20 30452613 30452786 3.60 0.58 -0.027 0.075 0.186 0.432 0.411 0.565 0.274 TGLA153 20 31240022 31240154 4.64 0.71 0.025 0.109 0.521 0.345 0.353 0.101 0.269 GHRp omS 20 31023202 31023306 3.12 0.43 0.006 0.114 0.581 0.426 0.446 0.726 0.268 BMS2361 20 34597279 34597368 5.10 0.72 0.019 0.125 0.698 0.329 0.351 0.045** 0.017** DIK4835 20 35915540 35916040 4.96 0.65 0.022 0.136 0.788 0.293 0.329 0.252 0.046 AGLA29 20 3842995 38843142 5.49 0.78 -0.006 0.087 0.202 0.363 0.412 0.000** 0.000** BMS117 20 40015465 40015564 3.88 0.67 -0.018 0.078 0.197 0.377 0.376 0.398 0.272 UMBTL78 20 40177064 40177157 4.22 0.58 -0.033 0.102 0.462 0.298 0.256 0.884 0.229 BM2113 2 88476 88616 5.44 0.79 -0.052 0.119 0.673 0.353 0.379 0.003** 0.005** INRA023 3 35576043 35576259 4.85 0.70 0.009 0.113 0.564 0.309 0.306 0.238 0.107 ETH10 5 55333999 55334220 4.57 0.67 0.002 0.134 0.789 0.432 0.446 0.049* 0.031* ETH152 5 NA NA 4.56 0.71 0.012 0.081 0.171 0.425 0.486 0.008** 0.020 ILSTS006 7 86555402 86555693 5.14 0.77 -0.007 0.076 0.110 0.331 0.351 0.032* 0.057 HEL9 8 NA NA 5.04 0.70 0.020 0.134 0.792 0.262 0.289 0.240 0.245 ETH225 9 8089454 8089601 5.02 0.71 0.013 0.113 0.560 0.410 0.478 0.009** 0.009** MM12 9 NA NA 7.76 0.67 0.017 0.123 0.671 0.312 0.347 0.244 0.112 ILSTS005 10 93304132 93304315 2.17 0.43 -0.026 0.083 0.356 0.686 0.664 0.358 0.390 CSRM60 10 70549981 70550081 7.03 0.72 0.011 0.073 0.094 0.405 0.418 0.046* 0.038* HEL13 11 NA NA 3.14 0.51 0.081 0.125 0.678 0.402 0.407 0.529 0.564 INRA032 11 49569411 49569592 3.81 0.62 -0.010 0.142 0.812 0.511 0.537 0.063 0.016 INRA037 11 70730695 70730819 4.54 0.58 0.030 0.129 0.717 0.266 0.243 0.830 0.462 INRA005 12 71751518 71751656 3.18 0.56 0.032 0.088 0.321 0.594 0.596 0.114 0.096 CSSM66 14 6128576 6128773 5.91 0.74 0.002 0.137 0.873 0.312 0.352 0.000** 0.003** HEL1 15 NA NA 3.99 0.67 0.020 0.072 0.138 0.468 0.445 0.119 0.155 SPS115 15 NA NA 5.40 0.58 0.039 0.096 0.416 0.478 0.482 0.228 0.146 INRA035 16 62926476 62926577 2.72 0.23 0.391 0.072 0.266 0.521 0.488 0.746 0.421 TGLA53 16 22214785 22214925 12.25 0.74 0.071 0.099 0.354 0.195 0.213 0.063 0.037 ETH185 17 36598852 36599086 8.31 0.68 0.039 0.146 0.877 0.336 0.303 0.186 0.196 INRA063 18 37562469 37562645 3.31 0.57 0.031 0.110 0.546 0.537 0.487 0.270 0.135 TGLA227 18 60360145 60360234 10.71 0.82 0.005 0.076 0.075 0.282 0.315 0.005** 0.012* ETH3 19 NA NA 4.44 0.65 0.009 0.135 0.787 0.407 0.406 0.073 0.139 HEL5 21 11850292 11850455 4.64 0.66 0.038 0.151 0.903 0.424 0.410 0.023* 0.104 TGLA122 21 50825795 50825936 11.36 0.74 0.007 0.069 0.065 0.210 0.213 0.538 0.152 Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 4 o 14 Fij iii iii ST(, ) exp( ) exp( ) =++ +++   1 whe e a i ,b j and g ij a e locus, popula ion and locus-by- popula ion pa ame e s, espec i ely [12]. In his s udy, he in e p e a ions o he po en ial ou lie s a e based on he locus e ec (a i ). Ou lie s om ou da a se we e iden i ied on he basis o he dis ibu ion ollowing [12]. Ra he han a ixed F ST as assumed in he abo e e- quen is me hod o [9], his BAYESFST es uses mo e in o ma ion om he aw da a and does no assume he same F ST o each popula ion [5,12]. Tes s o de ec loci unde selec ion o pai wise popula ions To es o addi ional e idence o selec ion, we used he combina ion o s a is ics lnRH, lnRV and lnRθ’in he popula ion pai wise compa isons. The p inciple behind hese es s is ha a iabili y a a neu al mic osa elli e locusisgi enbyθ=4N e μ,whe eN e is he e ec i e popula ion size and μis he mu a ion a e. A locus linked o a bene icial mu a ion will ha e a smalle e ec- i e popula ion size and consequen ly a educ ion in a iabili y below neu al expec a ions. The ela i e a - iance in a iabili y, lnRθ, can be assessed ins ead by es i- ma ing he ela i e a iance in epea numbe , lnRV, o he e ozygosi y, lnRH, o loci be ween popula ions. The lnRV was calcula ed using he equa ion lnRV = ln (V pop1 /V pop2 )whe eV pop1 and V pop2 a e he a iance in epea numbe o popula ion 1 and popula ion 2, espec i ely [10]. The lnRH es is based on he calcula- ion o he loga i hm o he a io o H o each locus o a pai o popula ions as ollows ln lnRH pop1 pop2 =− ⎛ ⎝ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟− − ⎛ ⎝ ⎜ ⎜ ⎞ ⎠ ⎟ ⎟− 1 11 1 11 2 2 H H whe e Hdeno es expec ed he e ozygosi y (see equa- ion 2 in [6]). In addi ion, we a emp ed o calcula e ln Rθby es ima ing θdi ec ly using a coalescence-based Bayesian Ma ko chain Mon e Ca lo simula ion app oach employing he MSVAR p og am [47]. The es s ha e been shown o be ela i ely insensi i e o mu a ion a e, de ia ion om he s epwise mu a ion model, demog aphic his o y o popula ion and sample size [16]. As sugges ed by [48], o de ec he mos ecen and s ong selec i e sweeps, he combina ion o lnRH andlnRVs a is icsisaspowe ulaslnRValone,bu using bo h s a is ics oge he lowe s he a e o alse posi i es by a ac o o 3 because he a iance in epea numbe and he he e ozygosi y o a popula ion measu e di e en aspec s o he a ia ion a a locus. Thus, com- bina ions o any wo o he h ee es s we e implemen- ed he e and signi icance o lnRH, lnRV and lnRθ’ o each compa ison was calcula ed acco ding o s anda d me hods [6,10,48]. These s a is ics a e gene ally no - mally dis ibu ed, and simula ions ha e con i med ha ou lie s (e.g. mo e han 1.96/2.58 s anda d de ia ions om he mean o 95%/99% con idence in e als, espec i ely) a e likely o be caused by selec ion [48]. The es s we e implemen ed o e e y pai wise compa i- son in ol ing na i e popula ions om di e en ai ca ego ies (Eas e n Finnca le, Wes e n Finnca le and No he n Finnca le s. Yaku ian, Is oben, Kholmogo y and Uk ianian G ey), i.e. 12 popula ion pai s o he ho n (polled/ho ned) ai . Tes s o de ec loci unde selec ion wi hin a popula ion The coalescence simula ion app oach using he De Sel 1.0 p og am [49] was used o de ec ou lie loci wi hin he Finnish na i e popula ions (Eas e n Finnca le, Wes e n Finnca le and No he n Finnca le). I has he ad an age o being able o ake in o accoun a wide ange o po en ial pa ame e s simul aneously and gi ing esul s ha a e obus ega ding he s a ing assump- ions. Fo each pai o popula ions (i,j), and o all loci, we calcula ed F i and F j (F i and F j a e he popula ion- speci ic di e gence; o de ails see [7,49]) and gene a ed he expec ed join dis ibu ion o F i and F j by pe o m- ing 10,000 coalescen simula ions. Thus, e e y locus all- ing ou side he esul ing con idence en elope can be seen as po en ially unde selec ion. The ollowing nui- sance pa ame e s we e used o gene a e null dis ibu- ions wi h simila numbe s o allelic s ages as in he Table 1 Summa y o he mic osa elli es and basic popula ion gene ic es ima es o he mic osa elli es (Con inued) HAUT24 22 45733839 45733962 7.09 0.70 0.025 0.143 0.861 0.406 0.424 0.004** 0.027* BM1818 23 35634770 35635033 4.03 0.63 0.019 0.102 0.458 0.538 0.486 0.144 0.013* HAUT27 26 26396836 26396987 8.85 0.61 0.126 0.103 0.453 0.376 0.396 0.083 0.003** BTA, Bos au us au osome; A R , allelic ichness; H E , expec ed he e ozygosi y, F IS , inb eeding coe icien , obse ed homozygosi y, F OBS , and expec ed homozygosi y, F EXP , NA, no a ailable; he p obabili ies o he Ewens-Wa e son es we e calcula ed based on homozygosi y (P H ) o Fishe s’s exac es (P E ); *, he signi icance le el o P< 0.05, **, he signi icance le el o P< 0.01; he genomic posi ions o he loci a e BLASTed agains STS o p ime sequence in ENSEMBL cow genome B au4.0 h p://www.ensembl.o g/Bos_ au us/In o/Index upda ed un il 11/02/2010 Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 5 o 14 obse ed da a se : mu a ion a es (in ini e allele model) μ=1×10 -2 ,1×10 -3 , and 1 × 10 -4 ; ances o popula ion size N e = 500, 5,000, and 50,000; imes since an assumed bo leneck e en T 0 = 50, 500, and 5,000 gene a ions; ime since di e gence = 50 and 500; and popula ion size be o e he spli N 0 = 50 and 500. In o de o de ec ou lie loci po en ially selec ed o he polled ai wi hin he h ee Finnish na i e ca le popula ions, he De Sel p og am was un o compa ison be ween he wo sub- popula ions ep esen ing he de ini ely polled (n=19) and ho ned (n= 19) animals, espec i ely. Resul s Gene ic di e si y and di e en ia ion A comple e lis o loci and hei a iabili y in he 10 ca - lepopula ionsa eshowninTable1.Theo e all gene ic di e en ia ion ac oss loci was 0.117 (F ST = 0.117, 95% CI 0.108 - 0.125). F ST alues o an indi i- dual locus a ied om 0.017 (SD = 0.011) a AGLA17 on BTA1 o 0.180 (SD = 0.057) a BMS2461 on BTA20. Mean popula ion di e en ia ions o loci on BTA1 and BTA20 we e 0.126 (F ST = 0.126, 95% CI 0.103 - 0.143) and 0.118 (F ST = 0.118, 95% CI 0.100 - 0.139), espec- i ely. Nei he o he alues indica ed signi ican di e - ence om he a e age o loci on o he ch omosomes (F ST = 0.114, 95% CI 0.104 - 0.124). Le els o a ia ion ac oss popula ions, including allelic ichness (A R ) and expec ed he e ozygosi y (H E ), we e in simila anges as o mic osa elli es on BTA1, BTA20 and o he au osomes, wi h he smalles a ia ions obse ed a AGLA17 (A R = 1.37, H E = 0.08). The highes H E o 0.79 was obse ed a BM2113 (BTA2) and he highes A R o 11.36 a TGLA122 (BTA21). Mos F IS alues we e posi i e and o some loci signi ican ly posi- i e. O he 13 nega i e F IS alues, se en occu ed o loci on BTA20, and wo o loci on BTA1. Loci on BTA1 and BTA20 did no show a signi ican educ ion o inc ease in mean F IS compa ed wi h he loci on o he au osomes (o he bo ine au osomes, mean F IS = 0.038; BTA1, mean F IS = 0.053, Mann-Whi ney es U= 118, P= 0.409; BTA20, mean F IS = 0.011, Mann-Whi ney es U= 273.5, P= 0.227). Gi en he ange o obse a- ions o F IS a an indi idual locus, he e we e no ma ked di e ence among he h ee classes o loci (BTA1, -0.083 - 0.190; BTA20, -0.074 - 0.113; o he BTAs, -0.052 - 0.391). Linkage disequilib ium The s eng h o pai wise linkage disequilib ium (LD) be ween ma ke s was es ima ed and he a e age D′ alue o pai wise syn enic ma ke s was 0.32 ac oss BTA1 and 0.28 ac oss BTA20, bo h o which a e signi i- can ly (P< 0.05) highe han o non-syn enic ma ke s (0.15; only he D′> 0.3 a e shown in Figu e 1). Figu e 1 also shows ma ices o LD signi icance le els o all pos- sible locus combina ions o he loci on BTA1 o BTA20 in hei ch omosomal o de . O he 120 pai wise com- pa isons o he 16 loci on BTA20, a o al o 22 (22/120, 18.3%) es s showed P alues below 0.05. Likewise, LD be ween ma ke s on BTA1 p o ided se en (7/21, 33.3%) signi ican obse a ions. Howe e , a subs an ially smalle p opo ion (34/1124, 3.0%) o signi ican (P< 0.05) pai s was ound be ween non-syn enic ma ke s. In gene al, signi ican ly highe le els o LD we e obse ed o syn e- nic ma ke s on BTA1 and BTA20 han ha o non- syn enic ma ke s. The e was no e idence o LD blocks on ei he o he ch omosomes. E idence o selec ion ac oss he popula ions The Ewens-Wa e son es enables de ec ion o de ia- ions om a neu al-equilib ium model as ei he a de i- ci o an excess o gene ic di e si y ela i e o he numbe o alleles a a locus (see [50]). When applying he es s o all he mic osa elli es, we de ec ed 13 loci (AGLA17,DIK5019,SOD1,AGLA29,BMS2361, BM2113,ETH10,ETH225,CSSM66,ETH152,TGLA227, HAUT24,andCSRM60) on 10 di e en ch omosomes exhibi ing signi ican p obabili ies o he Ewens-Wa e - son es based on bo h homozygosi y (P H )andFishe ’s exac es (P E ) (see Table 1). O he 13 loci, one (AGLA17) exhibi ed a signi ican (P< 0.05) de ici o he e ozygosi y and all he o he 12 loci exhibi ed a sig- ni ican (P< 0.05) excess in gene ic di e si y ela i e o he expec ed alues; hese pa e ns a e consis en wi h di ec ional and balancing selec ion, espec i ely. The 12 loci gene a ed a e age P alues signi ican ly (S uden ’s es : PH=0.020, = -5.65, P< 0.0001; PE=0.014, = -5.69, P< 0.0001) below han he expec ed median alue o 0.5. Howe e , a e age P alues o 0.313 o P H ( = -4.63, P> 0.1) and 0.232 o P E ( = -8.69, P>0.1) we e obse ed in he emaining 38 loci which we e no unde selec ion. The obse a ion p o ided u he e i- dence ha selec ion a ec ed gene ic di e si y a he mic osa elli es unde selec ion. The esul s o he analyses wi h he FDIST2 p og am a e p esen ed in Table 1 and Figu e 2a. This summa y- s a is ic me hod, based on simula ed and obse ed F ST alues, iden i ied ou loci (SOD1,BMS2461,DIK5019 and AGLA17) as ou lie s showing oo p in s o selec ion in he analyses, including all 10 popula ions, a he 5% signi icance le el. O he ou signi ican loci, h ee (SOD1,BMS2461 and DIK4519) wi h highe F ST alues indica ed a sign o di ec ional selec ion and one locus (AGLA17) appea ing in he lowe ail o he F ST dis i- bu ion sugges ed a signa u e po en ially a ec ed by bal- ancing selec ion (Figu e 2a). In he Bayesian F ST - es (Figu e 2b), which was based on a hie a chical eg es- sion model, h ee loci (HEL5,DIK4591and SOD1)we e Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 6 o 14 de ec ed as being di ec ionally selec ed and wo (AGLA17 and TGLA227) as unde balancing selec ion. O e all, ac oss all he popula ions, wo loci, AGLA17 and SOD1, exhibi ed he s onges e idence o selec ion wi h all h ee s a is ical app oaches, which p o ided good suppo o hei s a us as ou lie s due o selec ion. Two loci (DIK5019 and TGLA227) exhibi ed signi ican depa u e om he neu al expec a ions in wo ou o he h ee selec ion es s. Fu he mo e, 12 loci (AGLA29, BMS2361,BM2113,ETH10,ETH225,CSSM66,ETH152, Figu e 1 De ailed iew o he ex en and signi icance o LD in he ca le popula ions using he Haplo iew 4.0 p og am. Numbe s in he blocks indica e he pe cen age o he LD me ic D’ alues > 0.3; shadings indica e Fishe ’s exac es signi icance le els: whi e, P> 0.05; ligh shading, P< 0.05. Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 7 o 14 HAUT24,CSRM60, BMS2461, HEL5 and DIK4591) can be ega ded as candida es a ec ed by selec ion, bu we e e ealed only in one o he h ee es s. In e es ingly, acco ding o ENSEMBL cow genome h p://www. ensembl.o g/Bos_ au us/In o/Index he signi ican locus AGLA17 unde balancing selec ion was abou 1.78 cM ups eam om he candida e locus o POLL, whe eas locus SOD1 unde di ec ing selec ion was loca ed abou 3.87 cM downs eam om he candida e locus. I should be no ed ha he F ST -based es s o selec ion a e p one o alse posi i es because o sensi i i y o demog aphic his o y [51], he e ozygosi y among loci in mu a ion a e [52] and locus-speci ic phenomena no ela ed o selec- ion [48]. Ne e heless, we expec he se o loci iden i- ied by F ST -based es s o be en iched o he ue posi i es in u he es s. Tes s o selec ion o pai wise popula ions Since each o he i e es s used abo e elies on some- wha di e en assump ions, loci ha a e epea edly ound o be ou side he ange expec ed o neu ali y a e ex emely good candida es o ma ke s unde selec- ion. Mo eo e , LD is known o be ex emely high o he six BTA1 mic osa elli es nea he candida e gene a ec ing he p esence o absence o ho ns in Bos au us, hus he egion unde selec ion is likely o be qui e wide. Despi e he possible p esence o a ew alse posi- i es, he ull se o se en loci (SOD1,BMS2461, DIK5019,HEL5,DIK4591,TGLA227 and AGLA17)was used o u he analyses. The lnRθme hods (lnRH, lnRV and lnRθ’) use he e ozygosi y o a iance di e - ence, a he han popula ion di e gence, o es o selec ion. Signi ican esul s o he lnRθ es s o selec- i e sweeps in ol e he wo loci (AGLA17 and SOD1) de ec ed by he Ewens-Wa e son es and he F ST -based es s o pai wise combina ions (n= 12) o h ee na i e Finnish ca le popula ions and ou old na i e popula- ions om Russia and Uk aine (Table 2). Signi ican esul s o selec i e sweeps a loci AGLA17 and SOD1 we e ob ained o 12 pai wise popula ion Figu e 2 Resul s o (A) he FDIST2 and (B) BAYESFST es s. The solid lines indica e he c i ical cu o o he P- alue a he 0.05 le el. Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 8 o 14 compa isons o each o he h ee di e en measu es o lnRθ(Table 2). O he pai wise compa isons, a o al o 28 and 26 signi ican (P< 0.05) o e y signi ican (P< 0.01) esul s we e obse ed a AGLA17 and SOD1, espec i ely, in he h ee es s. Bo h loci (AGLA17 and SOD1) appea ed in all h ee di e en measu es o lnRθ o eigh o mo e compa isons (Table 2), ha is, lnRθ (lnRH, lnRV and lnRθ’) alues de ia ing by mo e han 1.96 s anda d de ia ions om he mean. Acco dingly, he pai wise compa isons be ween ei he o Eas e n Finnca le and Wes e n Finnca le and popula ions o Yaku ian, Kholmogo y and Uk ainian G ey we e signi i- can o all h ee es ima o s. All he compa isons be ween popula ions yielded a leas wo signi ican esul s o he h ee es ima o s. In o al, 54 (75% 54/72) signi ican compa isons in ol ed AGLA17 o SOD1 in he compa isons be ween Finnish na i e popula ions (No he n Finnca le, Eas e n Finnca le and Wes e n Finnca le) s. he na i e popula ions om Russia and Uk aine (Is oben, Uk ainian G ey, Kholmogo y and Yaku ian Ca le), which sugges ed ha selec i e sweeps had aken place in he Finnish na i e popula ions. Tes s o selec ion wi hin he Finnish na i e popula ions The coalescen simula ion, which was based on a popula- ion spli model [49], was pe o med wi h he De Sel p o- g am wi hin he Finnish na i e popula ions wi h e y simila demog aphical backg ounds (Eas e n Finnca le, No he n Finnca le and Wes e n Finnca le). Among he six BTA1 mic osa elli es a ound he candida e loci, all a e polymo phic in he h ee popula ions in ol ed in he pai wise-subpopula ion compa ison. In he pai wise com- pa ison be ween de ini ely polled (n= 19) and ho ned (n= 19) ca le, loci AGLA17 and SOD1 we e signi ican ly ou side he 99% con idence in e al (Figu e 3), while locus DIK4591 ell sligh ly ou side he 95% con idence en elope in he h ee compa isons, which a e hus con- side ed as alse posi i es, i.e., he locus was de ec ed as an ou lie because o he 5% ype I e o . The ou lie beha- io o loci AGLA17 and SOD1 was deemed o be he esul o s ong local e ec s o hi chhiking selec ion. Discussion In his s udy, besides 28 mic osa elli es on o he ca le au osomes used as a e e ence se o ma ke s, se en mic osa elli es on BTA1 and 16 on BTA20 a ound candi- da e loci we e sc eened o he oo p in s o selec ion among 10 ca le popula ions wi h di e gen ho n o p o- duc ion ai s. Ac oss di e en s a is ical analyses, a highly di e gen pa e n o gene ic di e en ia ion and la ge di e ences in le els o a iabili y we e e ealed a he loci SOD1 and AGLA17 among popula ions, which was inconsis en wi h neu al expec a ions. The esul s indica ed di e gen ‘selec i e sweeps’a AGLA17 and SOD1, p obably caused by selec ion o he closely-linked candida e loci o he ho ned/polled ai , e.g. he POLL gene. E idence o selec ion o mic osa elli es su ounding he POLL gene Because e ealing ou lie loci in genome scans cu en ly depends on s a is ical es s, one o he main conce ns is o highligh uly signi ican loci while minimizing he de ec ion o alse posi i es [44]. Using a mul ilocus scan o di e en ia ion based on mic osa elli e da a, we com- pa ed h ee di e en me hods ha aimed a de ec ing ou lie s om simula ed neu al expec a ions: 1) he Ewens-Wa e son me hod [44,45], 2) he FDIST2 me hod [9], and 3) a BAYESFST me hod [12]. Ou lie s we e iden i ied o 15 loci using a 5% h eshold, which was obus ac oss me hods o wo loci (SOD1 and AGLA17). The locus SOD1 p esen ed a highe Table 2 Es ima es o lnRV, lnRH and lnRθ’ o he pai wise compa isons Pai wise compa ison lnRV lnRH lnRθ’ AGLA17 SOD1 AGLA17 SOD1 AGLA17 SOD1 Eas e n Finnca le - Is oben * * n.s. n.s. * n.s. Eas e n Finnca le - Yaku ian * ** * ** ** * Eas e n Finnca le - Uk ainian G ey ** ** * * ** * Eas e n Finnca le - Kholmogo y * ** * * * * Wes e n Finnca le - Is oben ** * ** ** * * Wes e n Finnca le - Yaku ian ** ** * * * ** Wes e n Finnca le - Uk ainian G ey * * ** * * * Wes e n Finnca le - Kholmogo y * * * * * ** No he n Finnca le - Is oben * n.s * n.s. n.s. * No he n Finnca le - Yaku ian * n.s. n.s. * n.s. n.s. No he n Finnca le - Uk ainian G ey ** * n.s. n.s. n.s. n.s. No he n Finnca le - Kholmogo y * n.s. n.s. * n.s. n.s. * Signi icance P< 0.05, ** P< 0.01, n.s., no signi ican Li e al.Gene ics Selec ion E olu ion 2010, 42:32 h p://www.gsejou nal.o g/con en /42/1/32 Page 9 o 14