A Robust Test for Assortative Mating

European Journal of Human Genetics (2000) 8, 119–124 © 2000 Macmillan Publishers Ltd All rights reserved 1018–4813/00 $15.00 <a href="/goto?url=http://www.nature.com/ejhg" target="_blank">www.nature.com/e</a><a href="/goto?url=http://www.nature.com/ejhg" target="_blank">j</a><a href="/goto?url=http://www.nature.com/ejhg" target="_blank">h</a><a href="/goto?url=http://www.nature.com/ejhg" target="_blank">g </a>ARTICLE A robust test for assortative mating Em m an uelle Geń in 1, Carole Ober2, Lowell Weitkam p3 an d Glen ys Th om son 1 21Department of Integrative Biology, University of California, Berkeley, CA, USA; Department of Human Genetics, 3University of Chicago, Chicago, IL, USA; Department of Psychiatry and Division of Genetics, University of Rochester Medical Center, Rochester, NY, USA Testing for random mating in human populations is difficult due to confounding factors such as ethnic preference and population stratification. With HLA, the high level of polymorphism is an additional problem since it is rare for couples to share the same haplotype. Focus on an ethnically homogeneous population, where levels of polymorphism at HLA loci are more limited, may provide the best situation in which to detect non-random mating. However, such populations are often genetic isolates where there may be inbreeding to an extent that is difficult to quantify and account for. We have developed a test for random mating at a multiallelic locus that is robust to stratification and inbreeding. This test relies on the availability of genotypic information from the parents of both spouses. The focus of the test is on families where there is allele sharing between the parents of both spouses, so that potential spouses could share an allele. Denoting the shared allele at the locus of interest by A, then under the assumption of random mating, heterozygous parents AX should transmit allele A equally as frequently as allele X to their offspring. When there is positive (negative) assortative mating, A will be transmitted more (less) often than X. The power of the test has been computed in a number of situations. Data on high resolution HLA haplotypes from the Hutterite population were reinvestigated by the proposed test. The test detects significant negative assortative mating when the parental origin of the shared haplotype is taken into account. European Journal of Human Genetics (2000) 8, 119–124. Keywords: assortative mating; HLA; Hutterites; power; robustness; inbreeding; admixture Introduction leukocyte an tigen s (HLA) is th e very h igh level of polym orph ism in th is region th at lowers th e expectation of fin din g spouses sh arin g an iden tical allele or h aplotype. It h as th erefore been suggested th at studies of m ate ch oice in h um an s sh ould focus on eth n ically h om ogen eous population s, wh ere levels of polym orph ism at HLA loci is m ore lim ited. However, such population s are often gen etic isolates in wh ich th ere m ay be con siderable in breedin g or, altern atively, avoidan ce of m atin g with close relatives to an exten t th at is difficult to quan tify an d accoun t for. Th e purpose of th is study is to propose a test for assortative m atin g th at is n ot sen sitive to stratification an d/or in breedin g. Th is test relies on th e availability of gen otypic in form ation for paren ts of both spouses. Th e power of th e test is com puted for differen t m odels of positive an d n egative assortative m atin g an d allowin g for differen t types of relation sh ips between th e spouses. Th e properties of th e test are com pared to th ose of on e com m on ly used test of assortative m atin g based on th e distribution of th e n um ber of alleles sh ared by spouses.4–6 Mate ch oice in an im al an d h um an population s h as been th e subject of m an y in vestigation s. Assortative m atin g h as lon g been recogn ized in h um an s for ch aracteristics such as stature an d con gen ital deafn ess,1 wh ich in volve both gen etic an d en viron m en tal factors. Studies in m ice h ave im plicated gen es from th e m ajor h istocom patibility com plex (MHC) in m ate ch oice; fem ale m ice gen erally prefer m atin g with m ales differen t from th em selves for gen es in th e H2 region .2 However, attem pts to test for an effect of HLA on m ate ch oice in h um an s h ave led to equivocal con clusion s.3–6 Con foun din g factors such as eth n ic preferen ce an d population stratification m ake testin g of ran dom m atin g in h um an population s difficult. An oth er problem specific to h um an Correspondence: Emmanuelle Geńin, INSERM U535, Baˆ timent INSERM Gregory Pincus, secteur marron, porte 15, 80 rue du Geńe´ral Leclerc, 94276 Le Kremlin Biceˆ tre Cedex. Tel: + 33 1 49 59 53 30; Fax: + 33 1 49 59 53 31; E-mail: [email protected] Received 7 June 1999; revised 7 September 1999; accepted 23 September 1999 Testing for assortative mating E Geńin et al 120 Recen tly, two studies in h um an isolates h ave been publish ed,5,6 Th e first, in th e Hutterite population , sh owed eviden ce of n egative assortative m atin g for HLA: spouses m atch ed less often for HLA h aplotypes th an expected.5 Th is effect was n ot foun d wh en each locus was con sidered in dividually, suggestin g a h aplotype effect.7 Th e secon d study, in South Am erin dian tribes, sh owed n o deviation from ran dom m atin g wh en an alyzin g each locus separately. In both an alyses, assortative m atin g was tested by com parin g th e observed distribution of alleles or h aplotypes sh ared by spouses to th e on e expected un der ran dom m atin g an d th e sign ifican ce was depen den t on som e h ypoth eses regardin g th e structure of th e population . In th is paper, we rean alyze th e Hutterite data of Ober et al5 with th e n ew test we h ave proposed, wh ich does n ot depen d on an y h ypoth eses regardin g th e population structure. wh ere in form ative paren ts h ave tran sm itted th e sh ared allele Ai to th eir offsprin g (on e of th e spouses in gen eration G1). Th e statistic M defin ed by: 4(T – N/2)2 M = Nis distributed as a ch i-square with on e degree of freedom un der H0. Th e test we propose is based on th is statistic an d will be called th e M test. Depen din g on th e structure of th e fam ily, th e n um ber of in form ative pairs an d in form ative paren ts m ay vary (see Table 1). Special atten tion is drawn to th e case of fam ilies wh ere on e paren t sh ares on e allele in com m on with both paren ts of h is ch ild’s in -laws, th at is fam ilies in wh ich th e paren ts of th e first spouse are AiAj, AkAl an d th e paren ts of th e secon d spouse are AiAr, AjAs. In th ese fam ilies, on ly th e paren ts of th e secon d spouse sh ould be con sidered as in form ative with th e sh ared allele bein g respectively Ai an d Aj. Th e con sequen ce is th at th e test requires at least th ree alleles be presen t in th e fam ily. Th erefore it can n ot be used with biallelic m arkers. In deed, with biallelic m arkers, h eterozygous paren ts can on ly sh are alleles with both paren ts of th eir ch ild’s in -laws. Material and methods Description of th e m atin g test Th e robust test we h ave developed relies on th e availability of gen otypic in form ation for th e paren ts (gen eration G0) of both spouses (gen eration G1), an d focuses on ly on fam ilies wh ere th e paren ts sh are an allele, called Ai, at th e test locus un der study (see Table 1). Th e pair of paren ts sh arin g a com m on allele is called an in form ative pair. Un der th e assum ption of ran dom m atin g at gen eration G1, paren ts th at are h eterozygous AiAj tran sm it allele Ai as often as allele Aj to th eir offsprin g. If th e sh ared allele Ai is tran sm itted m ore often , it m ean s th at in gen eration G1, th ere was positive assortative m atin g with respect to th e locus un der study or a locus in lin kage disequilibrium with th at locus (see discussion ). Con versely, if Ai is tran sm itted less often th an Aj, it m ean s th at in gen eration G1, th ere was n egative assortative m atin g at th at locus. Statistical properties of th e M test com pared with th e IBS test Th e statistical properties of th e M test (robustn ess an d power) h ave been studied by sim ulation s. A com parison is perform ed with th e classically used iden tical by state (IBS) test.4–6 Description of the IBS test Th e IBS test is also a ch i-square test th at com pares th e observed distribution of spouses sh arin g 0, 1 or 2 IBS alleles to th e distribution expected un der ran dom m atin g (P0, P1, P2). To avoid an y assum ption regardin g th e m atin g pattern in th e paren tal gen eration , Hedrick8 proposed estim atin g P0, P1 an d P2 from th e data by Let N be th e n um ber of h eterozygous paren ts AiAj in in form ative pairs (we will refer to th ese paren ts as in form ative paren ts in th e followin g). Let T be th e n um ber of cases Table 1 Families considered in the analysis No. of informative pairs per family No. of informative <ul style="display: flex;"><li style="flex:1">parents </li><li style="flex:1">Type </li><li style="flex:1">Parents of 1st spouse </li><li style="flex:1">Parents of 2nd spouse </li></ul><ul style="display: flex;"><li style="flex:1">1</li><li style="flex:1">AA A A </li><li style="flex:1">AAm A A </li><li style="flex:1">i ≠ j, k, l </li></ul>m, r, s ≠ i, k, l i ≠ k, l 112 214<ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">j</li><li style="flex:1">k</li><li style="flex:1">l</li><li style="flex:1">i</li><li style="flex:1">r</li><li style="flex:1">s</li></ul>23<ul style="display: flex;"><li style="flex:1">AA A A </li><li style="flex:1">AAm A A </li></ul><ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">l</li><li style="flex:1">i</li><li style="flex:1">r</li><li style="flex:1">s</li></ul>m, r, s ≠ i, k, l i ≠ j, k, l l ≠ k <ul style="display: flex;"><li style="flex:1">AA A A </li><li style="flex:1">AAm A A </li></ul><ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">j</li><li style="flex:1">k</li><li style="flex:1">l</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">r</li></ul>m, r ≠ i, j, k, l i ≠ k, l 456<ul style="display: flex;"><li style="flex:1">AA A A </li><li style="flex:1">AAm A A </li><li style="flex:1">2</li></ul>22 322<ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">l</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">r</li></ul>l ≠ k m, r ≠ i, k, l i ≠ k, m <ul style="display: flex;"><li style="flex:1">AA A A </li><li style="flex:1">AAm A A </li></ul><ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">k</li><li style="flex:1">i</li><li style="flex:1">k</li><li style="flex:1">r</li></ul>m ≠ k r ≠ i, j, k i ≠ j, k, l m, r ≠ i, j, k, l <ul style="display: flex;"><li style="flex:1">AA A A </li><li style="flex:1">AAm AA </li></ul><ul style="display: flex;"><li style="flex:1">i</li><li style="flex:1">j</li><li style="flex:1">k</li><li style="flex:1">l</li><li style="flex:1">i</li><li style="flex:1">j</li><li style="flex:1">r</li></ul>European Journal of Human Genetics Testing for assortative mating E Geńin et al 121 usin g th e observed fem ale an d m ale gen otypic frequen cies in m atin g pairs from th e population . Th ese param eters clearly h ave an in fluen ce: 1. on th e probability of recruitin g fam ilies with in form ative pairs, an d 2. on th e power of th e M test to detect assortative m atin g.

A Robust Test for Assortative Mating

Details

Download

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

Support