A novel common variant in DCST2 is associated with length in early life and height in adulthood

A no el common a ian in DCST2 is associa ed wi h leng h in ea ly li e and heigh in adul hood Ral J.P. an de Valk1,2,3, { , Eskil K eine -Mølle 5, { , Ma jolein N. Kooijman1,2,3, { ,Mo `nica Guxens6,7,8, { , E angelia S e giakouli9, { , Annika Sa ¨a ¨ 12, Jona han P. B ad ield13, F ank Gelle 15, M. Geo ey Hayes16,17, Diana L. Cousmine 18, An je Ko ¨ ne 20, Elisabe h Thie ing21,22, John A. Cu in25, Ronny Myh e26, Ville Huika i28, Raimo Jo o31, Ma jan Ke kho 33,34, Nicole M. Wa ing on37,38, Niina Pi ka ¨nen39, Ioanna N alla41,42, Momoko Ho ikoshi43,44, Rii a Veijola45, Rachel M. F ea hy47, Yik-Ying Teo48,49,50, Sheila J. Ba on51, Da id M. E ans9,38, John P. Kemp9,38, Bea e S Pou cain9,10,11, Susan M. Ring9,10, Geo ge Da ey Smi h9, Anna Be gs o ¨m12, Inge Kull53,54, Hakon Hakona son13,55,14, F ank D. Men ch13, Hans Bisgaa d5, Bo Chawes5, Jakob S okholm5, Johannes Waage5, Pa ick E iksen5, As id Se els ed5, Mads Melbye15,56, Ea ly Gene ics and Li ecou se Epidemiology (EAGLE) Conso ium, Co nelia M. an Duijn1, Ca olina Medina-Gomez1,3,4, Albe Ho man1,3, Johan C. de Jongs e2,3, H. Rob Taal1,2, And e ´G. Ui e linden1,3,4, Gene ic In es iga ion o AN h opome ic T ai s (GIANT) Conso ium, Lo en L. A ms ong16,17, Johan E iksson18, Aa no Palo ie18,57,59,58, Ma iona Bus aman e6,7,8,61, Xa ie Es i ill7,8,61,62, Juan R. Gonzalez6,7,8, Sab ina Llop7,63, Wieland Kiess20, Anubha Mahajan43, Claudia Flexede 22, Ca la M.T. Tiesle 21,22, Cla e S. Mu ay25, Angela Simpson25, Pe Magnus27, Ve ena Sengpiel64, Anna-Liisa Ha ikainen29, Si kka Keinanen- Kiukaanniemi28, Alexand a Lewin65, Alexessande Da Sil a Cou o Al es65, Alexand a I. Blakemo e66, Jessica L. Bux on66, Ma ika Kaakinen28,65,30, Alina Rod iguez65,67, Syl ain Sebe 28, Ma ja Vaa asmaki46, Timo Lakka31,68,69, Vi pi Lindi31, Ul ike Geh ing70, Di kje S. Pos ma34,35, Wei Ang37, John P. Newnham37, Leo-Pekka Lyy ika ¨inen71,72, Ka ja Pahkala39,38, Olli T. Rai aka i39,74, Kalliope Panou sopoulou76, Ele he ia Zeggini76, Do e I. Boomsma77,78,79, Ma ia G oen-Blokhuis77,78,79, Jo ma Ilonen40,32, Lude F anke80, Joel N. Hi schho n81,60,82, Tune H. Pe s81,60,83, Liming Liang89, Jinyan Huang89,85, Be hold Hoche 86,87,88, Mikael Knip19,89,90, Seang-Mei Saw48,91,92, John W. Holloway52, E ikMele ´n12,54, S uanF.A. G an 13,55,14, Bja ke Feens a15, William L. Lowe16,17,Elisabe hWide ´n18, Elena Se geye 20, Ha ald G alle 23,24, Adnan Cus o ic25, Bo Jacobsson26,64, Ma jo-Rii a Ja elin28,65,30,93,94, Mus a a A alay31, Ge a d H. Koppelman34,36, C aig E. Pennell37, Ha i Niinikoski39,75, Geo ge V. Dedoussis42, Ma k I. Mcca hy43,44,95, Timo hy M. F ayling47, Jo di Sunye 6,7,8,62, { , Nicholas J. Timpson9, { , Fe nando Ri adenei a1,3,4, { , Klaus Bønnelykke5, { and Vincen W.V. Jaddoe1,2,3, { ,∗ o he Ea ly G ow h Gene ics (EGG) Conso ium † These au ho s ha e con ibu ed equally o his wo k. ‡ The au ho s wish i o be known ha , in hei opinion, J.S., N.J.T., F.R., K.B. and V.W.V.J. should be ega ded as join Lead Senio Au ho s; hese au ho s join ly di ec ed his wo k. ∗ To whom co espondence should be add essed a : Gene a ion R S udy G oup, Depa men o Epidemiology, E asmus Medical Cen e , Sophia’s Child en’s Hospi al, Pos bus 2060, 3000 CB Ro e dam, The Ne he lands. Tel: +31 107043405; Fax: +31 10 4089382; Email: [email p o ec ed] #The Au ho 2014. Published by Ox o d Uni e si y P ess. 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/4.0/),whichpe mi sun es ic ed euse,dis ibu ion,and ep oduc ioninanymedium,p o ided heo iginalwo kisp ope ly ci ed. Human Molecula Gene ics, 2015, Vol. 24, No. 4 1155–1168 doi:10.1093/hmg/ddu510 Ad ance Access published on Oc obe 3, 2014 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om 1 Depa men o Epidemiology, 2 Depa men o Paedia ics, 3 The Gene a ion R S udy G oup, 4 Depa men o In e nal Medicine, E asmus Medical Cen e , Ro e dam, The Ne he lands, 5 Copenhagen P ospec i e S udies on As hma in Childhood, Facul y o Heal h Sciences, Uni e si y o Copenhagen & Danish Pedia ic As hma Cen e , Copenhagen Uni e si y Hospi al, Gen o e, Denma k, 6 Cen e o Resea ch in En i onmen al Epidemiology (CREAL), Ba celona, Spain, 7 CIBER Epidemiologı ´a y Salud Pu ´blica (CIBERESP), Spain, 8 Pompeu Fab a Uni e si y (UPF), Ba celona, Ca alonia, Spain, 9 MRC In eg a i e Epidemiology Uni , 10 A on Longi udinal S udy o Pa en s and Child en (ALSPAC), School o Social and Communi y Medicine, 11 School o O al and Den al Sciences, Uni e si y o B is ol, B is ol, UK, 12 Ins i u e o En i onmen al Medicine, Ka olinska Ins i u e , S ockholm, Sweden, 13 Cen e o Applied Genomics, Ab amson Resea ch Cen e , 14 Di ision o Human Gene ics, The Child en’s Hospi al o Philadelphia, Philadelphia, PA 19104, USA, 15 Depa men o Epidemiology Resea ch, S a ens Se um Ins i u , Copenhagen, Denma k, 16 Di ision o Endoc inology, Me abolism and Molecula Medicine, 17 No hwes e n Uni e si y Feinbe g School o Medicine, Chicago, IL 60611, USA, 18 Ins i u e o Molecula Medicine Finland, 19 Diabe es and Obesi y Resea ch P og am, Uni e si y o Helsinki, Helsinki, Finland, 20 Cen e o Pedia ic Resea ch, Uni e si y Hospi al Cen e Leipzig, Uni e si y o Leipzig, Leipzig, Ge many, 21 Di ision o Me abolic and Nu i ional Medicine, D . on Haune Child en’s Hospi al, Uni e si y o Munich Medical Cen e , Munich, Ge many, 22 Ins i u e o Epidemiology I, 23 Ins i u e o Epidemiology II, 24 Resea ch Uni o Molecula Epidemiology, Helmhol z Zen um Mu ¨nchen – Ge man Resea ch Cen e o En i onmen al Heal h, Neuhe be g, Ge many, 25 Cen e o Respi a o y Medicine and Alle gy, Ins i u e o In lamma ion and Repai , Uni e si y o Manches e and Uni e si y Hospi al o Sou h Manches e , Manches e Academic Heal h Sciences Cen e, Manches e , UK, 26 Di ision Epidemiology, Depa men Genes and En i onmen , 27 Di ision Epidemiology, No wegian Ins i u e o Public Heal h, Oslo, No way, 28 Ins i u e o Heal h Sciences, 29 Ins i u e o Clinical Medicine/Obs e ics and Gynecology, 30 Biocen e Oulu, Uni e si y o Oulu, Oulu, Finland, 31 Ins i u e o Biomedicine, Physiology, 32 Depa men o Clinical Mic obiology, Uni e si y o Eas e n Finland, Kuopio, Finland, 33 Depa men o Epidemiology, 34 G oningen Resea ch Ins i u e o As hma and COPD, 35 Depa men o Pulmonology, 36 Bea ix Child en’s Hospi al, Pedia ic Pulmonology and Pedia ic Alle gy, Uni e si y o G oningen, Uni e si y Medical Cen e G oningen, G oningen, The Ne he lands, 37 School o Women’s and In an s’ Heal h, The Uni e si y o Wes e n Aus alia, Pe h, Aus alia, 38 Uni e si y o Queensland Diaman ina Ins i u e, T ansla ional Resea ch Ins i u e, B isbane, Queensland, Aus alia, 39 Resea ch Cen e o Applied and P e en i e Ca dio ascula Medicine, 40 Immunogene ics Labo a o y, Uni e si y o Tu ku, Tu ku, Finland, 41 Depa men o Heal h Sciences, Uni e si y o Leices e , Leices e LE1 7RH, UK, 42 Depa men o Nu i ion and Die e ics, Ha okopio Uni e si y o A hens, A hens 11527, G eece, 43 Wellcome T us Cen e o Human Gene ics, Uni e si y o Ox o d, Ox o d OX3 7BN, UK, 44 Ox o d Cen e o Diabe es, Endoc inology and Me abolism, Uni e si y o Ox o d, Chu chill Hospi al, Ox o d OX3 7LJ, UK, 45 Depa men o Pedia ics, Medical Resea ch Cen e , 46 Depa men o Obs e ics and Gynecology and MRC Oulu, Oulu Uni e si y Hospi al and Uni e si y o Oulu, Oulu, Finland, 47 Uni e si y o Exe e Medical School, Royal De on and Exe e Hospi al, Ba ack Road,Exe e EX2 5DW, UK, 48 Saw Swee Hock School o Public Heal h, 49 Li e Science Ins i u e, Na ional Uni e si y o Singapo e, Singapo e, 50 Genome Ins i u e o Singapo e, Agency o Science, Technology and Resea ch, 51 MRC Li ecou se Epidemiology Uni , 52 Human Gene ics and Genomic Medicine, Human De elopmen & Heal h, Facul y o Medicine, Uni e si y o Sou hamp on, UK, 53 Depa men o Clinical Science and Educa ion, So ¨de sjukhuse , S ockholm, Sweden, 54 Sachs’ Child en’s Hospi al, S ockholm, Sweden, 55 Depa men o Pedia ics, Pe elman School o Medicine, Uni e si y o Pennsyl ania, Philadelphia, PA, USA, 56 Depa men o Medicine, S an o d School o Medicine, S an o d, USA, 57 Analy ic and T ansla ional Gene ics Uni , Depa men o Medicine, 58 Psychia ic & Neu ode elopmen al Gene ics Uni , Depa men o Psychia y, Massachuse s Gene al Hospi al, Bos on, MA, USA, 59 P og am in Medical and Popula ion Gene ics, 60 Medical and Popula ion Gene ics P og am, B oad Ins i u e o MIT and Ha a d, Camb idge, MA, USA, 61 Cen e o Genomic Regula ion (CRG), Ba celona, Spain, 62 IMIM (Hospi al del Ma Medical Resea ch Ins i u e), Ba celona, Spain, 63 Founda ion o he P omo ion o Heal h and Biomedical Resea ch in he Valencian Region, FISABIO-Public Heal h, Valencia, Spain, 64 Depa men Obs e ics and Gynecology, Sahlg enska Academy, Sahlg enska Uni e si y Hospi al, Go henbu g, Sweden, 65 Depa men o Epidemiology and Bios a is ics, School o Public Heal h, Impe ial College London, MRC Heal h P o ec ion Agency (HPE) Cen e o En i onmen and Heal h, 66 Sec ion o In es iga i e Medicine, Di ision o Diabe es, Endoc inology, and Me abolism, Facul y o Medicine, Impe ial College, London W12 0NN, UK, 67 Depa men o Psychology, Mid Sweden Uni e si y, O ¨s e sund, Sweden, 68 Kuopio Resea ch Ins i u e o Exe cise Medicine, Kuopio, Finland, 69 Depa men o Clinical Physiology and Nuclea Medicine, Kuopio Uni e si y Hospi al, Kuopio, Finland, 70 Ins i u e o Risk Assessmen 1156 Human Molecula Gene ics, 2015, Vol. 24, No. 4 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om Sciences, U ech Uni e si y, U ech , The Ne he lands, 71 Depa men o Clinical Chemis y, Fimlab Labo a o ies, Tampe e, Finland, 72 Depa men o Clinical Chemis y, Uni e si y o Tampe e School o Medicine, Tampe e, Finland, 73 Spo s and Exe cise Medicine Uni , Depa men o Physical Ac i i y and Heal h, Paa o Nu mi Cen e, Tu ku, Finland, 74 Depa men o Clinical Physiology and Nuclea Medicine, 75 Depa men o Pedia ics, Tu ku Uni e si y Hospi al, Tu ku, Finland, 76 Wellcome T us Sange Ins i u e, The Mo gan Building, Wellcome T us Genome Campus, Hinx on, Camb idgeshi e CB10 1HH, UK, 77 Depa men o Biological Psychology, VU Uni e si y, Ams e dam, The Ne he lands, 78 EMGO Ins i u e o Heal h and Ca e Resea ch, Ams e dam, The Ne he lands, 79 Neu oscience Campus Ams e dam, The Ne he lands, 80 Depa men o Gene ics, Uni e si y o G oningen, Uni e si y Medical Cen e G oningen, The Ne he lands, 81 Di ision o Endoc inology and Cen e o Basic and T ansla ional Obesi y Resea ch, Bos on Child en’s Hospi al, USA, 82 Depa men o Gene ics, Ha a d Medical School, USA, 83 Cen e o Biological Sequence Analysis, Depa men o Sys ems Biology, Technical Uni e si y o Denma k, Denma k, 84 Depa men o Bios a is ics and Epidemiology, Ha a d School o Public Heal h, Bos on, USA, 85 Shanghai Ins i u e o Hema ology, Rui Jin Hospi al A ilia ed wi h Shanghai Jiao Tong Uni e si y School o Medicine, Shanghai, China, 86 Ins i u e o Nu i ional Science, Uni e si y o Po sdam, Ge many, 87 The Fi s A ilia ed Hospi al o Jinan Uni e si y, Guangzhou 510630, China, 88 Cen e o Ca dio ascula Resea ch/Ins i u e o Pha macology, Cha i e ´, Be lin, Ge many, 89 Depa men o Pedia ics, Tampe e Uni e si y Hospi al, Tampe e, Finland, 90 Child en’s Hospi al, Uni e si y o Helsinki and Helsinki Uni e si y Cen al Hospi al, Helsinki, Finland, 91 Singapo e Eye Resea ch Ins i u e, Singapo e, 92 Duke-NUS G adua e Medical School, Singapo e, 93 Uni o P ima y Ca e, Oulu Uni e si y Hospi al, Kajaanin ie 50, P.O.Box 20, FI-90220, Oulu 90029 OYS, Finland, 94 Depa men o Child en and Young People and Families, Na ional Ins i u e o Heal h and Wel a e, Aapis ie 1, Box 310, Oulu FI-90101, Finland and 95 Ox o d NIHR Biomedical Resea ch Cen e, Chu chill Hospi al, Ox o d OX3 7LJ, UK Recei ed July 3, 2014; Re ised and Accep ed Sep embe 29, 2014 Common gene ic a ian s ha e been iden i ied o adul heigh , bu no much is known abou he gene ics o skele al g ow h in ea ly li e. To iden i y common gene ic a ian s ha in luence e al skele al g ow h, we me a- analyzed 22 genome-wide associa ion s udies (S age 1; N528 459). We iden i ied se en independen op single nucleo ide polymo phisms (SNPs) (P<1310 26 ) o bi h leng h, o which h ee we e no el and ou we e in o nea loci known o be associa ed wi h adul heigh (LCORL,PTCH1,GPR126 and HMGA2). The h ee no el SNPs we e ollowed-up in nine eplica ion s udies (S age 2; N511 995), wi h s905938 in DC- STAMP domain con aining 2 (DCST2) genome-wide signi ican ly associa ed wi h bi h leng h in a join analysis (S ages 1 12; b 50.046, SE 50.008, P52.46 310 28 , explained a iance 50.05%). Rs905938 was also asso- cia ed wi h in an leng h (N528 228; P55.54 310 24 ) and adul heigh (N5127 513; P51.45 310 25 ). DCST2 is a DC-STAMP-like p o ein amily membe and DC-STAMP is an os eoclas cell- usion egula o . Polygenic sco es based on 180 SNPs p e iously associa ed wi h human adul s a u e explained 0.13% o a iance in bi h leng h. The same SNPs explained 2.95% o he a iance o in an leng h. O he 180 known adul heigh loci, 11 we e genome-wide signi ican ly associa ed wi h in an leng h (SF3B4,LCORL,SPAG17,C6o 173, PTCH1,GDF5,ZNFX1,HHIP,ACAN,HLA locus and HMGA2). This s udy highligh s ha common a ia ion in DCST2 in luences a ia ion in ea ly g ow h and adul heigh . INTRODUCTION Fe al and in ancy leng h g ow h a e impo an measu es o de- elopmen in ea ly li e. Ea ly leng h g ow h seems o be asso- cia ed wi h heigh in adul hood (1). I has been shown ha e al and in an g ow h a e independen ly associa ed wi h highe isks o ca dio ascula disease, ype 2 diabe es and many o he complex diseases. P e ious indings sugges ed gene ic links be ween e al g ow h and me abolism (2,3). Howe e , hese s udies mainly ocused on bi h weigh as ea ly g ow h measu e. Skele al g ow h is a di e en measu e o de elopmen in ea ly li e. Skele al g ow h du ing e al li e and in ancy is a complex ai wi h he i abili y es ima es o 26–72% (4). Al hough co ela ed wi h each o he , e al, in an and adul skele al g ow h may be in luenced by di e en gene ic ac o s. Many common gene ic a ian s ha e been iden- i ied o adul heigh (5), bu no much is known abou he gen- e ics o skele al g ow h in ea ly li e. Al hough, se e al a e gene ic de ec s wi h la ge e ec s on leng h a bi h and du ing in ancy ha e been ound (6,7), common gene ic a ian s ha in- luence no mal a ia ion in bi h and in an leng h ha e no ye been iden i ied. The e o e, we aimed o iden i y common gene ic a ian s in luencing ea ly leng h g ow h, also in pe spec- i e o hei e ec on adul s a u e. Human Molecula Gene ics, 2015, Vol. 24, No. 4 1157 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om RESULTS To iden i y common gene ic a ian s associa ed wi h bi h leng h, we examined 2 201 971 million di ec ly geno yped and impu ed SNPs wi h bi h leng h in 22 independen disco e y s udies wi h genome-wide associa ion (GWA) o Me abochip da a (S age 1; N¼28 459; Fig. 1). Bi h leng h was measu ed using s anda dized p ocedu es (Supplemen a y Ma e ial, Tables S1 and S2). S udies wi h sel - epo ed measu emen s we e excluded a p io i. Bi h leng h was s anda dized using g ow h analyze (h p://www.g ow hanalyse .o g), ans o ming bi h leng h in o sex- and age-adjus ed s anda d de ia ion sco es (SDS). We used he No h-Eu opean 1991 e e ence panel o compa e esul s be ween s udies. We applied linea eg ession be ween numbe o alleles o dosages ob ained om impu a ions and s anda dized bi h leng h ( ull de ails in Ma e ials and Me hods). Gene iden i ica ion In he disco e y phase (S age 1), we ound se en independen op SNPs wi h sugges i e e idence o associa ion (P,1×10 26 ) wi h bi h leng h (Supplemen a y Ma e ial, Figs. S1 and S2, QQ- and Manha an plo ). Fou SNPs mapped o loci al eady known o be associa ed wi h adul heigh (Supplemen a y Ma e - ial, Table S3, LCORL,PTCH1,GPR126 and HMGA2)(5). The 3 SNPs e lec ing po en ially no el associa ions we e aken o - wa d in nine independen eplica ion s udies (S age 2; N¼11 995; Fig. 1). Only one o he h ee SNPs displayed signi ican e idence o eplica ion in S age 2 and eached genome-wide signi icance in he join analysis (S ages 1 +2; P,5×10 28 ; Table 1). This no el associa ion a ose om SNP s905938, mapping o ch omosome 1q22 in DC-STAMP domain con ain- ing 2 (DCST2) (Fig. 2, egional associa ion plo ). Each C allele [mino allele equency (MAF) ¼0.24] o s905938 was asso- cia ed wi h an inc ease (s anda dized) o 0.046 SDS in bi h leng h (s anda d e o ¼0.008, P¼2.46 ×10 28 ; explained a iance ¼0.05%). The genome-wide signi ican ly associa ed SNP showed low deg ee o he e ogenei y be ween he disco e y s udies (P¼0.93, I 2 ¼0%). Figu e 3shows he o es plo o he associa ions be ween s905938[C] and bi h leng h ac oss all s udies. O he sugges i e loci in he disco e y analysis a e shown in Supplemen a y Ma e ial, Table S3 (P,1×10 25 ). Summa y s a is ics o all SNPs a e a ailable a h p://egg- conso ium.o g. Func ional analyses We assessed common a ian s wi h dele e ious unc ional impli- ca ions in linkage disequilib ium (LD, 2 .0.80) wi h s905938 using HaploReg (8). The e we e no non-synonymous a ian s in LD wi h s905938. We ound h ee pu a i e unc ional in onic a ian s in high LD wi h s905938. De ails a e depic ed in Sup- plemen a y Ma e ial, Table S4. Subsequen ly, we assessed whe he a ian s in he iden i ied locus we e in ol ed in he Figu e 1 S udy design. 1158 Human Molecula Gene ics, 2015, Vol. 24, No. 4 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om egula ion o messenge RNA exp ession (eQTLs) in genome- wide exp ession da ase s o lymphoblas oid cell lines (LCLs, N¼1830) (9,10). We ound cis eQTLs [ alse disco e y a e (FDR) ,1% accoun o all SNP-p obe pai s ha we e wi hin 1 Mb o each o he ) o ansc ip s o PBXIP1,GBA and ADAM15. Ye , s905938 and he cis eQTL SNPs we e no in pe ec LD ( 2 ,0.80, Supplemen a y Ma e ial, Table S5). The e o e, we canno exclude ha mul iple independen e ec s a ise om he same egion o associa ion. DCST2 and g ow h pheno ypes We es ed he associa ions o s905938[C] wi h ‘ e al g ow h’ measu es in he 1s , 2nd and 3 d imes e o p egnancy in he Table 1 Summa y s a is ics o he h ee no el SNPs a P,1×10 26 in he disco e y analysis and he eplica ion ollow-up esul s Ma ke MAF b SE PnI 2 He P Disco e y (S age 1) s905938[C] a 1q22 (DCST2) 0.24 0.050 0.010 2.59 ×10 27 28 327 0.0 0.930 s12545524[G] a 8q22.1 (nea GDF6) 0.14 0.078 0.014 1.54 ×10 28 22 170 6.6 0.376 s11037473[A] a 11p11.2 (nea es genes TTC17-HSD17B12) 0.06 20.109 0.021 2.17 ×10 27 22 259 0.0 0.735 Replica ion (S age 2) s905938[C] a 1q22 (DCST2) 0.23 0.035 0.015 1.99 ×10 22 11 908 – – s12545524[G] a 8q22.1 (nea GDF6) 0.11 20.012 0.017 4.67 ×10 21 17 614 – – s11037473[A] a 11p11.2 (nea es genes TTC17-HSD17B12) 0.08 20.035 0.020 8.06 ×10 22 17 606 – – Disco e y + eplica ion (S ages 1 +2) s905938[C] a 1q22 (DCST2) 0.24 0.046 0.008 2.46 ×10 28 40 235 – – s12545524[G] a 8q22.1 (nea GDF6) 0.13 0.042 0.011 9.08 ×10 25 39 784 – – s11037473[A] a 11p11.2 (nea es genes TTC17-HSD17B12) 0.07 20.069 0.014 1.49 ×10 26 39 865 – – SNPs ma ke s a e iden i ied acco ding o hei s anda d s numbe s (NCBI build 36). Independen no el SNPs wi h a s ong sugges i ee ec in he disco e y analysis on bi h leng h a e shown (P,1×10 26 ). SNPs in loci ha a e known o be associa ed wi h adul heigh we e excluded o eplica ion e o s (adul heigh loci: LCORL,PTCH1,GPR126 and HMGA2). MAF, mino allele equency; SE, s anda d e o . b e lec s di e ences in s anda dized bi h leng h pe mino allele. P alues a e ob ained om linea eg ession o each SNP agains s anda dized bi h leng h adjus ed o sex and ges a ional age. We included bo hGWA and me abochip coho s in ou disco e y analysis, s905938 is on he me abochip, and s12545524 and s11037473 a e no , his explains he di e ences in numbe s (n). De i ed inconsis ency s a is ic I 2 and He P alues e lec he e ogenei y ac oss disco e y s udies wi h he use o Coch an’s Q es s. Figu e 2 Regionalassocia ion plo o 1q22 in he 22 bi h leng hdisco e y s udies(N¼28 459).SNPsa e plo edwi h hei P alues (as 2log 10 alues;le y-axis) as a unc ion o genomic posi ion (x-axis). Es ima ed ecombina ion a es ( igh y-axis) aken om HapMap a e plo ed o e lec he local LD-s uc u e a ound he op associa edSNP(‘whi eopendiamond’)and heco ela edp oxies(‘ci cles’acco ding oablack- o-g ayscale om 2 ¼0 o 1).Thejoin analysis P alueo disco e y and eplica ion s udies is epo ed wi h he ‘whi e squa e’ (N¼40 235). Human Molecula Gene ics, 2015, Vol. 24, No. 4 1159 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om Gene a ion R S udy (N¼5756) (11), in an leng h a 1 yea o age ( ange 6–18 mon hs; N¼28 228) in he Ea ly G ow h Gene ics (EGG) conso ium (12), and adul heigh in he Gene ic In es iga ion o An h opome ic T ai s (GIANT) con- so ium (N¼127 513) (5). Rs905938[C] was no associa ed wi h ‘ e al g ow h’ measu es, bu was associa ed wi h in an leng h and adul heigh (P,0.05; Table 2). Known adul heigh loci in ela ion o bi h and in an leng h We also explo ed whe he common gene ic a ian s known o be associa ed wi h adul heigh (5) in luenced bi h leng h a ia ion. We ound ha 17 ou o 180 known adul heigh loci we e asso- cia ed wi h bi h leng h (FDR ,5%, Supplemen a y Ma e ial, Table S6; Fig. 4,QQ-plo o 180 SNPs and bi h leng h). We hen calcula ed a heigh -inc easing-alleles sco e o he 180 known heigh loci (5) o p edic bi h leng h in he Gene a ion R S udy (N¼2085; Fig. 5). The sco e composed o a ian s associa ed wi h adul heigh explained 0.13% o he a iance in bi h leng h (P¼0.1), in con as o he 10% o he pheno ypic a ia ion in adul heigh epo ed in he o iginal manusc ip (5). To e alua e whe he di e en common gene ic a ian s in lu- enced bo h bi h and in an leng h, we es ed 2 193 675 million SNPs o associa ion wi h in an leng h in almos he same se o samples used o he analysis o bi h leng h (19 s udies, N¼28 238; Supplemen a y Ma e ial, Table S7). We iden i ied genome-wide signi ican associa ions a 11 gene ic loci (Supple- men a y Ma e ial, Figs S3 and S4, QQ- and Manha an plo ), which all a e known o be associa ed wi h adul heigh (Table 3, SNPs in o nea SF3B4,LCORL,SPAG17,C6o 173, PTCH1,GDF5,ZNFX1,HHIP,ACAN,HLA locus and HMGA2) (5,13). In addi ion, we ound ha a ian s in 58 o he adul heigh loci we e associa ed wi h in an leng h a an FDR o 5% (Supplemen a y Ma e ial, Table S8; Fig. 4,QQ-plo o 180 SNPs and in an leng h). Nex , we es ed in he Gene a ion R S udy (N¼2385) how much o he pheno ypic a iance in in an leng h was explained by he sco e composed o heigh - inc easing-alleles. Va ian s om he 180 known adul heigh loci oge he explained 2.95% o he a iance in in an leng h (P¼3.10 ×10 217 , Fig. 5). DEPICT analysis o bi h and in an leng h Finally, we used a pa hway analysis ool called DEPICT (Pe s e al., unpublished da a) o p io i ize genes a associa ed egions, sea ch o econs i u ed gene se s ha we e en iched in genes nea associa ed a ian s, and iden i y issue and cell ypes in which genes om loci associa ed wi h bi h and in an leng h we e highly exp essed ( ull de ails in Ma e ials and Me hods). Fo bo h ai s, we used independen SNPs ( 2 ,0.05) asso- cia ed a P,1×10 25 , om 21 bi h leng h and 44 in an leng h loci. The e we e no pa hways signi ican ly o e ep e- sen ed in he bi h leng h esul s. In con as , o in an leng h DEPICT signi ican ly p io i ized nine genes which we e o e e- p esen ed (FDR ,5%, Supplemen a y Ma e ial, Table S9), including h ee known Mendelian human s a u e genes (ACAN, GDF5 and PTCH1) as well as se e al ele an econs i u ed Figu e 3 Fo es plo o he associa ions be ween s905938[C] and bi h leng h. ∗Replica ion s udies. The ‘black diamond’ indica es he o e all e ec size and he con idence in e al o he 31 s udies. Table 2 Associa ions o s905938[C] in DCST2 ela ed o bi h leng h wi h ‘ e al g ow h’ measu es, in an leng h and adul heigh b SE P Gene a ion R: e al g ow h (N¼5756) Fi s imes e C own- ump leng h (n¼1126) 0.003 0.045 0.952 Second imes e Femu leng h (n¼5361) 20.035 0.023 0.129 Thi d imes e Femu leng h (n¼5532) 20.015 0.022 0.490 EGG: in an leng h In an leng h a 1 yea o age (N¼28 228) 0.035 0.010 5.54 ×10 24 GIANT: adul heigh Adul heigh (N¼127 513) 0.024 0.006 1.45 ×10 25 s905938 C-allele wi h a genome-wide signi ican e ec on bi h leng h is shown (P,5×10 28 ) in ela ion o ‘ e al g ow h’ measu es, in an leng h and adul heigh . SE, s anda d e o . b e lec s di e ence in s anda d de ia ion sco es pe mino allele. 1160 Human Molecula Gene ics, 2015, Vol. 24, No. 4 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om gene se s (e.g. abno mal s e num ossi ica ion, egula ion o os eoblas p oli e a ion and WNT signaling, Supplemen a y Ma e ial, Table S10). The e was no signi ican en ichmen o pa icula issue o cell ypes o any o he wo ai s. DISCUSSION In he p esen s udy we iden i ied one p e iously unknown locus ( s905938 in DCST2 a 1q22) o be associa ed wi h bi h leng h a a genome-wide signi ican le el. This common gene ic a ian was also associa ed wi h in an leng h and adul heigh . I was no possible o iden i y eQTLs o ansc ip s o DCST2 in he MRCA and MRCE da abases, as he e we e no p obes a ailable (9). Also, he e was no signi ican eQTL o DCST2 in immo alized LCLs (10). Howe e , DCST2 is a DC-STAMP- like p o ein amily membe and DC-STAMP is an impo an egula o o os eoclas cell- usion in bone homeos asis (14–16). The ansc ip s o PBXIP1,GBA and ADAM15 we e in weak LD wi h ou lead SNP s905938. The PBXIP1 p o ein is known o egula e es ogen ecep o unc ions (17). Mu a ions in he GBA gene cause Gauche disease, and s ong associa ions wi h Pa kinson’s disease and demen ia wi h Lewy bodies ha e been desc ibed (18–21). ADAM15 is p ominen ly exp essed in os eo- blas s and o a lesse ex en in os eoclas s (22). A s udy in mice showed ha ADAM15 is equi ed o no mal skele al homeos a- sis and ha i s absence causes inc eased nuclea ansloca ion o b-ca enin in os eoblas s leading o inc eased os eoblas p oli e - a ion and unc ion, which esul s in highe abecula and co ical bone mass (23). The 1q22 locus is a complex egion ha bo ing mul iple in e es ing genes ha could a ec bi h leng h. We em- phasize ha we could no speci ically pinpoin he causal gene(s) as ou lead SNP ( s905938) was no in pe ec LD wi h ou cis eQTL SNPs. Al hough, he e is some o e lap be ween adul heigh loci and bi h leng h, which is illus a ed by 17 sha ed loci, he gene ic a chi ec u e o adul heigh seems mo e simila o he gene ic a chi ec u e o in an leng h han bi h leng h [58 sha ed loci o in an leng h, based on conse a i e s a is ical me hod (FDR)]. One poin o conside a ion o he in e p e a ion o ou indings is he po en ial o measu emen e o o bi h leng h (24). This may lead o less powe o de ec no el gene ic a ian s as s anda d e o s o SNPs could be inc eased. The es i- ma e o he isk-allele sco e slope o Figu e 5is no in luenced by measu emen e o and he di e ences in he slopes sugges ha bi h and in an leng h a e in luenced by dis inc gene ic a ian s. We ound ha he SNP e ec s o bi h leng h o 137 o he 180 es ablished heigh loci we e in he same di ec ion as epo ed in he GIANT pape (5) (Supplemen a y Ma e ial, Table S6; p obabili y o success ¼0.761, P¼6.25 ×10 213 ). One hund ed six y- wo o he 180 loci we e in he same di ec ion o in an leng h (Supplemen a y Ma e ial, Table S8; p obabili y o success ¼0.900, P¼2.20 ×10 216 ). Fou SNPs associa ed wi h bi h leng h (P,1×10 25 ) a e in o nea loci known o be associa ed wi h bi h weigh (LCORL, HMGA2, ADCY5 and ADRB1). LCORL is associa ed wi h bi h weigh , bi h leng h, in an leng h and adul heigh , bu we could no ind an ob ious link be ween he gene and adul -onse diseases. HMGA2 is associa ed wi h ao ic oo size (25), ype 2 diabe es (26), and many o he ai s like oo h de elopmen , head ci cum e ence and b ain s uc u e (12,27). ADCY5 is also associa ed wi h ype 2 diabe es and ADRB1 wi h adul blood p essu e (2,3). These indings highligh gene ic links be ween e al g ow h and me abolism (2,3,26). As we ound o e lap be ween gene ic a ian s o bi h weigh and bi h leng h, we looked-up he e ec o s905938 in DCST2 on bi h weigh in a p e ious EGG s udy (3).Rs905938 was associa ed wi h bi h weigh , bu weake as compa ed wi h bi h leng h ( b ¼0.035 SDS, SE ¼0.010, P¼2.35 ×10 24 ,N¼26 558). In conclusion, in he p esen s udy we iden i ied one no el locus ( s905938 in DCST2 a 1q22) associa ed wi h bi h leng h a a genome-wide signi ican le el. This common Figu e 4 QQ-plo s o he 180 known adul heigh SNPs wi h bi h and in an leng h. QQ-plo o he 180 known adul heigh SNPs in associa ion wi h bi h leng h (uppe panel) in 22 s udies (N¼28 459) and wi h in an leng h (lowe panel) in 19 s udies (N¼28 238). The black do s ep esen obse ed P alues and he diagonal lines ep esen he expec edP alues unde he null dis ibu ion. Human Molecula Gene ics, 2015, Vol. 24, No. 4 1161 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om gene ic a ian was also associa ed wi h in an leng h and adul heigh , wi h dec easing magni ude o he associa ions in la e li e (0.046 SDS o bi h leng h, 0.035 SDS o in an leng h and 0.024 SDS o adul heigh ). To ou knowledge, no pheno- ype has been p e iously associa ed wi h he DCST2 gene and while he gene is exp essed in os eoclas s, i s unc ion should be u he s udied. MATERIALS AND METHODS S age 1: disco e y genome-wide associa ion analyses o bi h leng h We combined 21 popula ion-based s udies wi h GWA o Me a- bochip da a and bi h leng h a ailable ( o al N¼28 459 indi i- duals). One o ou disco e y coho s had wo independen sub-samples wi hin hei s udy leading o a o al o 22 independ- en GWA/Me abochip sub-samples o ou analysis: one sub- sample om heA onLongi udinalS udyo Pa en sandChild en (ALSPAC, GWA, n¼4816); Child en, Alle gy, Milieu, S ock- holm, Epidemiology [Swedish] (BAMSE, GWA, n¼423); Chil- d en’s Hospi al O Philadelphia (CHOP, GWA, n¼432); Copenhagen S udy on As hma in Childhood 2000 (COPSAC- 2000, GWA, n¼348); Copenhagen S udy on As hma in Child- hood Regis y (COPSAC-Regis y, GWA, n¼1111); Danish Na ional Bi h Coho (DNBC, GWA, n¼932); Gene a ion R S udy (Gene a ion R, GWA, n¼2085); Hype glycemia and Ad e se P egnancy Ou comes s udy (HAPO, GWA, n¼1325); Helsinki Bi h Coho S udy (HBCS, GWA, n¼1572); In ancia y Medio Ambien e (INMA, GWA, n¼848); Leipzig Childhood Obesi y coho (LEIPZIG, Me bochip, n¼607); Li es yle Immune Sys em Alle gy s udy (LISA, GWA, n¼552); Figu e 5 Heigh -inc easing-alleles sco e o known adul heigh SNPs p edic ing bi h and in an leng h. Gene ic isk-allele sco es (sum o heigh -inc easing alleles weigh ed by known e ec on adul heigh (5) ans o med o s anda d de ia ion Z-sco es) in he Gene a ion R s udy plo ed agains leng h adjus ed o sex and age. The dis ibu ion o he gene ic isk-allele sco e is depic ed as ba s. (A) Mean bi h leng h plo ed agains he gene ic sco e (N¼2085). (B) Mean in an leng h plo ed agains he gene ic sco e (N¼2385). Table 3 Summa y s a is ics o he ele en known adul heigh SNPs in associa ion wi h in an leng h a P,5×10 28 Ma ke MAF b SE PnI 2 He P s7536458[G] a 1p12 (SPAG17) 0.25 20.064 0.010 9.61 ×10 211 28234 0.0 0.403 s11205303[C] a 1q21.2 (SF3B4) 0.34 0.087 0.011 1.79 ×10 216 26559 0.0 0.864 s1380294[T] a 4p15.31 (LCORL) 0.15 20.108 0.014 2.54 ×10 214 23079 13.7 0.184 s1812175[A] a 4q28-q32(HHIP) 0.18 20.068 0.011 2.33 ×10 29 28227 0.0 0.398 s592229[G] a (HLA locus) 0.43 0.048 0.009 2.22 ×10 28 28223 0.6 0.326 s9385399[T] a 6q22.32 (C6o 173) 0.46 0.055 0.009 1.68 ×10 210 28224 0.0 0.943 s1984119[C] a 9q22.3 (PTCH1) 0.26 20.063 0.010 1.77 ×10 210 28197 0.0 0.490 s7970350[T] a 12q15 (HMGA2) 0.49 20.047 0.009 2.90 ×10 28 28226 0.0 0.426 s2280470[A] a 15q26.1 (ACAN) 0.36 0.053 0.009 6.43 ×10 29 27443 0.0 0.436 s143384[G] a 20q11.2 (GDF5) 0.44 0.058 0.009 2.87 ×10 210 28232 0.0 0.996 s1567865[T] a 20q13.13 (ZNFX1) 0.21 0.063 0.010 1.10 ×10 29 28229 22.5 0.104 SNPs ma ke s a e iden i ied acco ding o hei s anda d s numbe s (NCBI build 36). The o al sample includes da a o 19 independen da ase s (N¼28 238). MAF, mino allele equency;SE, s anda de o . b e lec sdi e encesins anda dized in an leng hpe mino allele. P aluesa eob ained om linea eg essiono each SNP agains s anda dized in an leng h adjus ed o sex and age. We included bo h GWA and me abochip coho s in ou disco e y analysis, his explains he di e ences in numbe s (n). De i ed inconsis ency s a is ic I 2 and He P alues e lec he e ogenei y ac oss disco e y s udies wi h he use o Coch an’s Q es s. 1162 Human Molecula Gene ics, 2015, Vol. 24, No. 4 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om Manches e As hma and Alle gy S udy (MAAS, GWA, n¼402); No wegian Mo he and Child Coho s udy (MOBA, GWA, n¼832); No he n Finland Bi h Coho s 1966 (NFBC66, GWA, n¼4642); No he n Finland Bi h Coho s 1986 (NFBC86, Me abochip, n¼4652); Physical Ac i i y and Nu i- ion in Child en s udy (PANIC, Me abochip, n¼319); wo sub- samples om he P e en ion and Incidence o As hma and Mi e Alle gy bi h coho s udy (PIAMA1, GWA, n¼283; PIAMA2, GWA, n¼195); The Wes e n Aus alian P egnancy Coho S udy (RAINE, GWA, n¼1272); Special Tu ku Co ona y Risk Fac o In e en ion P ojec (STRIP, Me abochip, n¼614); and TEENs o A ica: Genes and En i onmen (TEENAGE, GWA, n¼197). While no sys ema ic pheno ypic di e ences we e obse ed be ween he sub-samples o he PIAMA bi h coho s udy, hey we e analyzed sepa a ely due o geno yping on di e en pla o ms and a di e en ime pe iods. Geno ypes wi hin each s udy we e ob ained using high- densi y SNP a ays and hen impu ed o 2.5 M HapMap SNPs (Phase II, elease 22; h p://hapmap.ncbi.nlm.nih.go /). The basic cha ac e is ics, exclusions applied ( o example, indi i- duals o non-Eu opean ances y, amily ela ed indi iduals), geno yping, quali y con ol and impu a ion me hods o each dis- co e y s udy a e p esen ed in Supplemen a y Ma e ial, Table S1. S a is ical analysis wi hin disco e y s udies In all s udies, bi h leng h was measu ed using s anda dized p o- cedu es.S udieswi hsel - epo edmeasu emen swe eexcluded a p io i. Bi h leng h was s anda dized using g ow h analyze (h p://www.g ow hanalyse .o g), ans o ming bi h leng h in o sex- and age-adjus ed SDS. We used he No h-Eu opean 1991 e e ence panel o compa e esul s be ween s udies. Mul- iple bi hs and wins we e excluded om all analyses. We applied linea eg ession be ween numbe o alleles o dosages ob ained om impu a ions and s anda dized bi h leng h. The GWA analysis pe s udy was pe o med using MaCH2q l (28), SNPTEST (29), PLINK (30) o P opABEL (31). The secu ed da a exchange and s o age we e acili a ed by he E asmus Medical Cen e , Depa men o In e nal Medicine (32). Me a-analysis o disco e y s udies P io o me a-analysis, SNPswi h aMAF ,0.01 andpoo ly impu ed SNPs [ 2ha ,0.3 (MaCH); p ope _in o ,0.4 (IMPUTE2); R2_BEALE ,0.4 (BEAGLE)] we e il e ed. Genomic con ol (GC) (33) was applied o adjus he s a is ics gene a ed wi hin each coho (see Supplemen a y Ma e ial, Table S1 o indi id- uals udy l alues). Fou ou o he wen y- wo sub-samples we e geno yped on Me abochips. These SNP-a ays we e en iched wi h ‘adul heigh SNPs’. No mal a ia ion in ea ly leng h g ow h seems o be associa ed wi h heigh in adul hood (1). The e o e, we assumed mo e ue-posi i e hi s in hese s udies and did no apply GC in hese s udies (GIANT e al., unpublished da a). De ails o any addi ional co ec ions o s udy speci ic popula ion s uc u e a e gi en in he Supplemen a y Ma e ial, Table S1. In e se a iance ixed-e ec s me a-analyses we e analyzed using METAL ( eleased 2010-08-01) (34) by wo me a-analys s in pa allel and blinded o ob ain iden ical esul s. A e he METAL me a-analysis, we il e ed SNPs wi h a MAF ,0.05 and SNPs ha we e no a ailable in a leas 12 sub- samples o a oid alse-posi i e indings. We used Coch an’s Q es and he de i ed inconsis ency s a is ic I 2 o assess e idence o be ween-s udy he e ogenei y o he e ec sizes. The me a-analysis esul s we e ob ained o a o al o 2 201 971 SNPs. SNPs ha c ossed he h eshold o P≤1×10 26 we e conside ed o ep esen s ong sugges i e e idence o associ- a ion wi h bi h leng h. SNPs ha we e al eady known o be asso- cia ed wi h adul heigh we e excluded o he eplica ionanalysis (5). The explained a iance o he op SNPs we e calcula ed inone o he la ges coho s, he Gene a ion R S udy (n¼2085). S age 2: eplica ion analysis o op bi h leng h SNPs In he disco e y phase, we ound se en independen SNPs wi h s ong sugges i e e idence o associa ion (P,1×10 26 ) wi h bi h leng h. Fou SNPs we e al eady known o be associa ed wi h adul heigh (5). These SNPs we e excluded o ollow- up analyses. The h ee emaining no el SNPs we e ollowed-up in eplica ion s udies. We included bo h GWA and Me abochip s udies in ou disco e y analysis. Rs905938 was on ou Me abo- chips, and s12545524 and s11037473 we e no . This esul s in di e ences in numbe s o ou op SNPs in he disco e y and ep- lica ion analyses. s905938 was aken o wa d in 9 independen eplica ion s udies (N¼11 995), s12545524 and s11037473 in 13 independen eplica ion s udies including he ou disco e y Me abochip s udies (N¼17 679). De ails o he eplica ion s udies a e p esen ed in Supplemen a y Ma e ial, Table S2. Wi hin he eplica ion s udies, we analyzed he associa ion be ween numbe o alleles and s anda dized bi h leng h. Com- bined e ec es ima es and he e ogenei y be ween coho s was calcula ed using ixed e ec s me a-analyses in R Ve sion 2.8.1 (The R ounda ion o S a is ical Compu ing, lib a y me a). Top SNPs ha c ossed he signi ican h eshold o P- eplica ion ≤0.05 and he widely accep ed genome-wide signi icance h eshold o P≤5×10 28 o all s udies combined we e consid- e ed o ep esen obus e idence o associa ionwi h bi hleng h. The ins i u ional e iew boa ds o human s udies app o ed he p o ocols and w i en consen was ob ained om he pa icipa - ing subjec s o hei ca egi e s i equi ed by he ins i u ional e iew boa d. DEPICT analysis We used he no el Da a-d i en Exp ession-P io i ized In eg a- ion o Complex T ai s (DEPICT) me hod (Pe s e al., unpub- lished da a). DEPICT is designed o sys ema ically iden i y he mos likely causal gene a a gi en locus, gene se s ha a e en iched in gene ic associa ions, and issues and cell ypes in which genes om associa ed loci a e highly exp essed. Fi s , DEPICT assigns genes o associa ed SNPs using LD 2 .0.5 dis ance o de ine locus bounda ies, me ges o e lapping loci and disca ds loci mapping wi hin he ex ended majo his ocom- pa ibili y complex egion (ch omosome 6, base pai s 25 000– 35 000). Nex , he DEPICT me hod p io i izes genes wi hin a gi en associa ed locus based on he genes’ unc ional simila i y o genes om o he associa ed loci. Genes ha a e highly simila o genes om o he loci ob ain low p io i iza ion P alues, and simula ed GWAS esul s a e used o adjus o gene leng h bias as well as o he po en ial con ounde s. The e can be se e al p io i ized genes in a gi en locus. Nex , DEPICT conduc s gene se en ichmen analysis by es ing whe he genes in associa ed loci en ich o econs i u ed e sions o known pa hways, gene Human Molecula Gene ics, 2015, Vol. 24, No. 4 1163 a Tampe e Uni e si y Lib a y. Depa men o Heal h Sciences on Sep embe 27, 2016h p://hmg.ox o djou nals.o g/Downloaded om