Meta-analysis of genome-wide association studies identifies six new Loci for serum calcium concentrations

 Insektenkunde

 2 views
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Description
Meta-analysis of genome-wide association studies identifies six new Loci for serum calcium concentrations
Share
Tags
Transcript
  ARTICLE Meta-Analysis of Genome-Wide Association StudiesIdentifies 1q22 as a Susceptibility Locusfor Intracerebral Hemorrhage Daniel Woo, 1,34, * Guido J. Falcone, 2,3,4,5,6,34 William J. Devan, 2,3,4,5 W. Mark Brown, 7 Alessandro Biffi, 2,3,4,5 Timothy D. Howard, 7 Christopher D. Anderson, 2,3,4,5 H. Bart Brouwers, 2,3,4,5 Valerie Valant, 2,3,4,5 Thomas W.K. Battey, 2,3,4,5 Farid Radmanesh, 2,3,4,5 Miriam R. Raffeld, 2,3,4,5 Sylvia Baedorf-Kassis, 2,3,4,5 Ranjan Deka, 8  Jessica G. Woo, 9 Lisa J. Martin, 10 Mary Haverbusch, 1 Charles J. Moomaw, 1 Guangyun Sun, 8  Joseph P. Broderick, 1 Matthew L. Flaherty, 1 Sharyl R. Martini, 1 Dawn O. Kleindorfer, 1 Brett Kissela, 1 Mary E. Comeau, 7  Jeremiasz M. Jagiella, 11 Helena Schmidt, 12 Paul Freudenberger, 12 Alexander Pichler, 13 Christian Enzinger, 13,14 Bjo¨rn M. Hansen, 15,16 Bo Norrving, 15,16  Jordi Jimenez-Conde, 17,18 Eva Giralt-Steinhauer, 17,18 Roberto Elosua, 17,18 Elisa Cuadrado-Godia, 17,18 Carolina Soriano, 17,18  Jaume Roquer, 17,18 Peter Kraft, 6 Alison M. Ayres, 4 Kristin Schwab, 4  Jacob L. McCauley, 19  Joanna Pera, 11 Andrzej Urbanik, 20 Natalia S. Rost, 2,3,4,5  Joshua N. Goldstein, 21 Anand Viswanathan, 4 Eva-Maria Sto¨gerer, 13 David L. Tirschwell, 22 Magdy Selim, 23 Devin L. Brown, 24 Scott L. Silliman, 25 Bradford B. Worrall, 26  James F. Meschia, 27 Chelsea S. Kidwell, 28  Joan Montaner, 29 Israel Fernandez-Cadenas, 29,30 Pilar Delgado, 29 Rainer Malik, 31,32 Martin Dichgans, 31,32 Steven M. Greenberg, 4 Peter M. Rothwell, 33 Arne Lindgren, 15,16 Agnieszka Slowik, 11 Reinhold Schmidt, 13 Carl D. Langefeld, 7,35 and Jonathan Rosand 2,3,4,5,35, *on behalf of the International Stroke Genetics Consortium Intracerebralhemorrhage(ICH)isthestrokesubtypewiththeworstprognosisandhasnoestablishedacutetreatment.ICHisclassifiedaslobar or nonlobarbased on the locationof ruptured blood vessels within the brain. These different locations also signal different under-lyingvascularpathologies.Heritability estimates indicatea substantialgeneticcontributiontorisk of ICHin both locations. We reportagenome-wide association study of this condition that meta-analyzed data from six studies that enrolled individuals of Europeanancestry. Case subjects were ascertained by neurologists blinded to genotype data and classified as lobar or nonlobar based on braincomputed tomography. ICH-free control subjects were sampled from ambulatory clinics or random digit dialing. Replication of signalsidentified in the discovery cohort with p  <  1  3  10  6 was pursued in an independent multiethnic sample utilizing both direct andgenome-wide genotyping. The discovery phase included a case cohort of 1,545 individuals (664 lobar and 881 nonlobar cases) and acontrol cohort of 1,481 individuals and identified two susceptibility loci: for lobar ICH, chromosomal region 12q21.1 (rs11179580,odds ratio [OR]  ¼  1.56, p  ¼  7.0  3  10  8 ); and for nonlobar ICH, chromosomal region 1q22 (rs2984613, OR   ¼  1.44, p  ¼  1.6  3  10  8 ). 1 Department of Neurology and Rehabilitation Medicine, University of Cincinnati College of Medicine, Cincinnati, OH 45267, USA;  2 Center for HumanGenetic Research, Massachusetts General Hospital, Boston, MA 02114, USA;  3 Division of Neurocritical Care and Emergency Neurology, Department of Neurology,MassachusettsGeneralHospital,Boston,MA02114,USA; 4 TheJ.PhilipKistlerStroke ResearchCenter,MassachusettsGeneralHospital,Boston,MA 02114, USA;  5 Program in Medical and Population Genetics, Broad Institute, Cambridge, MA 02141, USA;  6 Department of Epidemiology, HarvardSchool of Public Health, Boston, MA 02115, USA;  7 Center for Public Health Genomics and Department of Biostatistical Sciences, Wake Forest University,Winston-Salem, NC 27157, USA;  8 Department of Environmental Health, University of Cincinnati College of Medicine, Cincinnati, OH 45267, USA;  9 Di-vision of Biostatistics and Epidemiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH 45229, USA;  10 Division of Human Genetics,Cincinnati Children’s Hospital Medical Center, Cincinnati, OH 45229, USA;  11 Department of Neurology, Jagiellonian University Medical College, Krakow31-008, Poland; 12 Institute ofMolecularBiology and Medical Biochemistry, Medical UniversityGraz, Graz 8010, Austria;  13 DepartmentofNeurology, Med-ical University of Graz, Graz 8036, Austria;  14 Division of Neuroradiology, Department of Radiology, Medical University of Graz, Graz 8036, Austria; 15 Department of Clinical Sciences Lund, Neurology, Lund University, Lund 221 85, Sweden;  16 Department of Neurology, Ska˚ne University Hospital,Lund 221 85, Sweden;  17 Department of Neurology, Neurovascular Research Unit, Institut Hospital del Mar d’Investigacions Me`diques, Universitat Auton-oma de Barcelona/DCEXS-UPF, Barcelona 08003, Spain;  18 Cardiovascular Epidemiology and Genetics Research Group, Institut Hospital del Mar d’Inves-tigacions Me`diques, Barcelona 08003, Spain;  19  John P. Hussman Institute for Human Genomics, University of Miami, Miller School of Medicine, Miami,FL 33136, USA;  20 Department of Radiology, Jagiellonian University Medical College, Krakow 31-008, Poland;  21 Department of Emergency Medicine, Mas-sachusetts General Hospital, Boston, MA 02114, USA;  22 Stroke Center, Harborview Medical Center, University of Washington, Seattle, WA 98104, USA; 23 Department of Neurology, Stroke Division, Beth Israel Deaconess Medical Center, Boston, MA 02215, USA;  24 Stroke Program, Department of Neurology,University of Michigan Health System, Ann Arbor, MI 48109, USA;  25 Department of Neurology, University of Florida College of Medicine, Jacksonville, FL32209,USA; 26 DepartmentofNeurologyandPublicHealthSciences,UniversityofVirginiaHealthSystem,Charlottesville,VA22908,USA; 27 Departmentof Neurology,MayoClinic,Jacksonville,FL32224,USA; 28 DepartmentofNeurology,UniversityofArizona,Tucson,AZ85724,USA; 29 NeurovascularResearchLaboratory and Neurovascular Unit, Institut de Recerca, Hospital Vall d’Hebron, Universitat Autonoma de Barcelona, Barcelona 08035, Spain;  30 StrokePharmacogenomics and Genetics, Fundacio´ Doce`ncia i Recerca Mu´tuaterrassa, Barcelona 08010, Spain;  31 Institute for Stroke and Dementia Research, Kli-nikum der Universita¨t Mu¨nchen, Ludwig-Maximilians-University, Munich 80539, Germany;  32 Munich Cluster for Systems Neurology (Synergy), Munich80539, Germany;  33 Stroke Prevention Research Unit, Nuffield Department of Clinical Neuroscience, University of Oxford, Oxford OX3 9DU, UK 34 These authors contributed equally to this work  35 These authors contributed equally to this work and are co-senior authors*Correspondence: daniel.woo@uc.edu (D.W.), jrosand@partners.org (J.R.) http://dx.doi.org/10.1016/j.ajhg.2014.02.012.  2014 by The American Society of Human Genetics. All rights reserved. The American Journal of Human Genetics  94 , 1–11, April 3, 2014  1 Please cite this article in press as: Woo et al., Meta-Analysis of Genome-Wide Association Studies Identifies 1q22 as a Susceptibility Locus forIntracerebral Hemorrhage, The American Journal of Human Genetics (2014), http://dx.doi.org/10.1016/j.ajhg.2014.02.012  The replication included a case cohort of 1,681 individuals (484 lobar and 1,194 nonlobar cases) and a control cohort of 2,261 individ-uals and corroborated the association for 1q22 (p  ¼  6.5 3 10  4 ; meta-analysis p  ¼  2.2 3 10  10 ) but not for 12q21.1 (p  ¼  0.55; meta-analysis p  ¼  2.6  3  10  5 ). These results demonstrate biological heterogeneity across ICH subtypes and highlight the importance of ascertaining ICH cases accordingly. Introduction Strokeisthemostdevastatingmanifestation ofcerebrovas-cular disease and the second leading cause of death andacquired disability worldwide. 1,2 Strokes are classified asischemic or hemorrhagic (85% and 15% in people of Euro-pean descent, respectively), and spontaneous intracerebralhemorrhage (ICH [MIM 614519]) is by far the most com-mon type of hemorrhagic stroke. 3 Largely a disease of theelderly, ICH occurs when cerebral vessels rupture andis associated with 40%–50% 3-month mortality as wellas sustained disability in more than half of survivors. 4,5 The incidence of ICH is expected to rise in coming yearsbecause of increases in life expectancy and widespreaduse of antithrombotic therapy in the elderly. 6,7 Althoughtreatment of hypertension partially reduces the risk of initial or recurrent ICH, 8 there are no established acutetreatments for this condition. Therefore, identification of biological pathways that could eventually be targeted bynovel therapeutic strategies is vital to reducing the healthcare burden associated with this disease.Histopathological observations demonstrate that theunderlying cerebral small vessel disease differs accordingtothelocationoftheICHwithinthebrain.LobarICHorig-inatesinthecerebralcortexorcortical-subcorticaljunctionand is most commonly associated with cerebral amyloidangiopathy. 9 Nonlobar ICH srcinates in deep structuresof cerebral hemispheres, brainstem, and cerebellum andtends to be associated with what is typically called hyper-tensive vasculopathy. 10 This heterogeneity in underlyingbiological processes leading to different ICH subtypeshas been corroborated by epidemiologic, 11 neuroimag-ing, 12 and genetic studies. 13,14 A large multicenter candi-date gene study undertaken by the International StrokeGenetics Consortium in the same populations utilized inthe present study established that the epsilon variants of   APOE  (MIM 107741), known to be risk factors for sporadiccerebral amyloid angiopathy, are associated at genome-wide significance levels specifically with lobar ICH. 13 Like-wise, it was subsequently found that the burden of risk alleles for high blood pressure associates specifically withnonlobar ICH. 14 Heritability estimates indicate that common geneticvariation plays a substantial role in risk of both ICH sub-types beyond  APOE  and blood-pressure-related variants. 15 Identification of these genetic contributors would have asignificant impact on the field of stroke, because it couldhelp uncover specific biological pathways hitherto unsus-pected to play a role in this condition that could betargeted by novel therapeutic strategies. In this studywe meta-analyzed data from six previously unpublishedgenome-wide association studies of ICH that enrolled sub-jects of European ancestry in the United States and Europeunder the auspices of the International Stroke GeneticsConsortium, with subsequent replication of identified sus-ceptibility loci in an independent study of ICH. Subjects and Methods Participating Studies Case and control subjects included in the discovery phase weresubjects of European ancestry aged  > 55 years in the Geneticsof Cerebral Hemorrhage with Anticoagulation 13 (GOCHA) study(multicenter study in the US) and aged  > 18 years in the Geneticand Environmental Risk Factors for Hemorrhagic Stroke 16 (GERFHS) studies I and II in Cincinnati, OH; Hospital del MarIntracerebral Hemorrhage 17 study and Vall d’Hebron HospitalICH 18 study in Barcelona, Spain; Jagiellonian University Hemor-rhagic Stroke Study 19 in Krakow, Poland; and the Lund StrokeRegister 20 study in Lund, Sweden. Because of their limited samplesizes, data from the four European studies (ESs) were analyzedtogether for the purposes of quality control, imputation, andassociation testing. Subjects Cases were ascertained across participating studies according topredefined standardized criteria. Spontaneous ICH was definedas a new and acute neurological deficit with compatible brain im-aging (computed tomography or magnetic resonance imaging)showing the presence of intraparenchymal bleeding. Accordingto standard research and clinical practice in the field, 3 ICH loca-tion was assigned based on admission images by neurologistswho were blinded to genotype data. ICH srcinating at the cere-bral cortex or cortical-subcortical junction (with or withoutinvolvement of subcortical white matter) was defined as lobar,and ICH srcinating at the thalamus, internal capsule, basalganglia, deep periventricular white matter, cerebellum, or brainstem was defined as nonlobar. Exclusion criteria included trauma,brain tumor, hemorrhagic transformation of ischemic stroke,vascular malformation, and any other cause of secondary ICH.Control subjects were ICH-free individuals enrolled from thesame population that gave rise to the case subjects at each partici-pating study site, aged  > 55 years (GOCHA) and  > 18 years(GERFHSandESs).Controlsubjectsweresampledbyrandomdigitdialing in GERFHS and from ambulatory clinics in the remainderof the studies.All studies were approved by the Institutional Review Board orethics committee at each participating site. Participants providedinformed consent; when subjects were not able to communicate,consent was obtained from their legal proxies. Genome-wide Genotyping and Quality Control Case and control subjects were genotyped with Affymetrix 6.0in GERFHS and with Illumina HumanHap610-Quad in GOCHAand ESs. Case and control subjects from each study were geno-typed side-by-side on the same plates with the exception of thereplication controls from the Cincinnati Control Cohort (CCC) 2  The American Journal of Human Genetics  94 , 1–11, April 3, 2014 Please cite this article in press as: Woo et al., Meta-Analysis of Genome-Wide Association Studies Identifies 1q22 as a Susceptibility Locus forIntracerebral Hemorrhage, The American Journal of Human Genetics (2014), http://dx.doi.org/10.1016/j.ajhg.2014.02.012  andGenomicControlCohort(GCC),whichweregenotypedfromseparate studies. Plate-to-plate variability was assessed by compar-ison of SNP call and error rates. Standardized prespecified quality-control procedures 21 were implemented separately in GOCHA,GERFHS, and ESs. These filters excluded SNPs with genotype callrate  < 0.95, significant differential missingness between case andcontrol subjects (p < 0.05), deviation from Hardy-Weinberg equi-librium (p < 1 3 10  6 ), or minor allele frequency (MAF) < 0.01. Atthe subject level, quality control excluded individuals with geno-type call rate of   < 95%; inconsistency between self-reported andgenotypic gender; an inferred first- or second-degree relative inthe sample identified on the basis of pairwise allele sharing esti-mates (estimated genome proportion shared identical by descent; p > 0.1875); and extreme genome-wide heterozygosity f statistic,defined as > 5 times its standard deviation. Population Stratification After quality-control procedures, principal-components anal-ysis 22 was implemented separately in GOCHA, GERFHS, andESs, incorporating genotype data from 1000 Genomes 23 popula-tions. Population outliers were identified and removed by visualinspection of principal component plots, and the first four prin-cipal components were subsequently included in regressionmodels fitted for association testing. Principal-component andidentity-by-descent analyses were performed via a pruned subsetof independent SNPs (61,325 SNPs in GOCHA, 95,013 inGERFHS, and 64,728 in ESs) to account for potential biases intro-duced by LD structure. Imputation After quality-control procedures and principal component anal-ysis, imputation was performed separately in GOCHA, GERFHS,andESsviaIMPUTE2v.2.2 24 and1000Genomes 23 integratedrefer-encepanels(PhaseIinterimreleaseinNCBIbuild37).Thenumberof SNPs that entered the imputation process were 525,752,795,240, and 532,149 for GOCHA, GERFHS, and ESs, respectively.Postimputation filters excluded imputed SNPs with MAF  < 0.01,IMPUTE2 information score  < 0.7, confidence score  < 0.9, andmissing estimates in association testing for 1 or more studies. Genome-wide Association Testing  Given the biological differences in ICH subtypes outlined in theintroduction and after a prespecified analysis plan, genome-wideassociation analyses were computed separately for all ICH (lobarand nonlobar combined), for lobar ICH, and for nonlobar ICH.These analyses were completed separately in GOCHA, GERFHS,and ESs via logistic regression, assuming additive genetic effects(1-degree-of-freedom additive trend test) and adjusting for age,gender, and principal components. In secondary analysis, associa-tion testing was carried out separately in each European study.Association p values obtained in GOCHA, GERFHS, and ESs weremeta-analyzed via the inverse normal method weighting bysample size as implemented in METAL, 25 and heterogeneity of pooled estimates was quantified by computing Cochrane’s Q and corresponding p and I 2 . After recent GWAS meta-analysis, 26 only SNPs with available estimates in all three data sets arereported.Quantile-quantileplotswereutilizedtoassesssystematicinflationinassociationresultsresultingfrompopulationstratifica-tion or other systematic causes of bias.Further analyses were undertaken to evaluate the presence of additional independent signals at each locus. Additional indepen-dent signals at each locus were evaluated by conditional testingcompleted by adding the dosages of the top SNP at each locus tologistic regression models. Pairwise linkage disequilibrium (LD)between SNPs was assessed and visualized with 1000 GenomesEnsemble-based genome browser. 27 Regional association plotswere constructed with LocusZoom software. 28 Replication Replication of associations with p  <  1  3  10  6 in the discoverymeta-analysis was pursued in case and control groups from theEthnic/Racial Variations of Intracerebral Hemorrhage (ERICH) 29 and GERFHS III 16 studies, in adult control group from theCCC, 16 and GCC. 16 Enrolled ICH cases and controls fromERICH were non-Hispanic whites, African Americans, and His-panic subjects (based on self-reported race and ethnicity) aged > 18 years from multiple study centers across the US. ICH casesubjects were ascertained with the same criteria utilized in thediscovery phase. ICH-free control subjects were sampled byrandom digit dialing from the population that gave rise to thecase cohort. The CCC and GCC cohorts are population-basedcohorts of ICH-free individuals from the Greater Cincinnatiarea. The CCC cohort was specifically enrolled to approximatethe age, sex, and race distribution of ICH subjects, and theGCC was enrolled to match the population and geographic dis-tribution of the metropolitan area. Genotyping was completedwith TaqMan assays (ERICH Study) or Affymetrix 6.0 (GERFHSIII and CCC-GCC). For the latter, preimputation quality-controlprocedures, imputation, and postimputation filters were imple-mented as described for the discovery phase. Association testingwas carried out by fitting logistic regression models and imple-menting a 1-degree-of-freedom trend test, assuming additiveeffects and including age and gender in all models. Meta-ana-lyses across discovery and replication proceeded as describedabove for the discovery phase. Replication results were consid-ered significant at p  <  0.05 and genome-wide significance wasdefined as p  <  5  3  10  8 . 30 Overlap with Common Variants Related to BloodPressure Given the well-established role of hypertension in causing ICH,especially of nonlobar type, we specifically assessed the role inICH of common genetic variants known to play a role in deter-mining blood pressure. SNPs reported to be related to blood pres-sure at p  <  1  3  10  5 were identified in the GWAS Catalog. 31 Queried traits included blood pressure; hypertension; anddiastolic, systolic, mean, and pulse blood pressures. Associationresults for these SNPs were identified for all, for lobar, and fornonlobar ICH. Overlap with DNase I Hypersensitivity Sites Thepositionsof SNPs in ICH-associated loci wereoverlapped withDNase I hotspot regions from the Encyclopedia of DNA Elements(ENCODE) Project that mark generalized chromatin accessibility,mapped for each of 125 diverse cell lines and tissues. 32 The genomic region of interest for each identified loci was definedbased on the genomic variants that were in linkage disequilib-rium with the top variant at each locus (defined as r 2 >  0.5). Inaddition, ICH-associated SNPs were analyzed for other overlapwith ENCODE data, including transcription factor motifs, viaRegulomeDB. 33 The American Journal of Human Genetics  94 , 1–11, April 3, 2014  3 Please cite this article in press as: Woo et al., Meta-Analysis of Genome-Wide Association Studies Identifies 1q22 as a Susceptibility Locus forIntracerebral Hemorrhage, The American Journal of Human Genetics (2014), http://dx.doi.org/10.1016/j.ajhg.2014.02.012  eQTL Analyses Variants within each identified susceptibility locus for ICH wereevaluated for gene expression in  cis  via publicly available re-sources. SNPs with p  <  1  3  10  5 were assessed in four publiclyavailable eQTL databases: SCAN (SNP and CNV AnnotationDatabase), the NCBI and Broad Institute GTEx (Genotype-TissueExpression) eQTL Browsers, 34 the Pritchard laboratory UChicagoeQTL browser, 35 and mRNA by SNP Browser v.1.0.1. 36 Geneexpression was assessed in a range of tissue and cell types,includingliver, brain,lymphoblastoid celllines,monocytes,fibro-blasts,andTcells.Asinpreviousreports, 37 wedefinedpotential cis eQTLsascandidateSNPsassociatedwithgeneexpressionmappingto a 1 Mb window around each locus with p < 1 3 10  3 . Results After excluding subjects based on quality-control proce-dures (n  ¼  23) and principal component analysis (n  ¼ 93), a case cohort of 1,545 subjects and a control cohortof 1,481 subjects were available for association testing inthe discovery analysis (mean age 67 [SD 10], female sex45%, Table 1). After preimputation quality-control proce-dures, imputation to 1000 Genomes reference panels,and postimputation quality-control filters, a total of 5,258,103SNPswereavailableforassociationtestingacrossall data sets included in the discovery sample.Susceptibility loci were identified for lobar and nonlobarICH, but not for all ICH (lobar and nonlobar combined)(Figures 1A–1C, Tables S1–S3 available online). The esti- mated inflation factors ( l  of 1.039, 1.016, and 1.038 forall, lobar, and nonlobar ICH, respectively) and quantile-quantile plots (Figure S1) indicated absence of inflationresulting from systematic bias caused by population sub-structure or other artifacts. Several SNPs on chromosomalregion 12q21.1, an intergenic region near  TRHDE  (MIM606950), were associated with lobar ICH, with peak associ-ation detected at rs11179580 (MAF ¼  0.24, per additionalmajor allele [C], odds ratio [OR] 1.56, 95% confidenceinterval [CI] 1.33–1.84; p ¼ 7.0 3 10  8 ; Q  ¼ 0.43, I 2 ¼ 0%;Figure 1B and Table 2). Similar results were obtained when analyzing each European study separately (Figure S2). Asimilar effect was observed for all ICH (per additional Callele, OR 1.36, 95% CI 1.21–1.54; p  ¼  5.4  3  10  8 ; Q   ¼ 0.07,I 2 ¼ 62%;Table2andFigure1A),butthecombination ofsimilareffectmagnitudes(despitedoublethesamplesize)and significant increase in heterogeneity suggests that theobserved association is driven by lobar ICH. rs11179580also showed some effect in nonlobar ICH (OR 1.25, 95%CI 1.09–1.42, p  ¼  0.002; Table 2), indicating that theaffected biological pathway could impact both types of ce-rebralhemorrhage.Neitheradjustmentforthemostsignif-icant SNPs at this locus (Figure S3A) nor haplotype testingidentified additional associations of interest.For nonlobarICH, a susceptibility locus was identifiedonchromosomal region 1q22, a region that contains  PMF1 (MIM609176) and  SLC25A44 (MIM 610824). The top-asso-ciated variant within this locus was the intronic SNPrs2984613 (MAF 0.31; per additional major allele [C] OR 1.44, 95% CI 1.27–1.64; p  ¼  1.6  3  10  8 ; Q   ¼  0.05, I 2 ¼ 66%; Table 2 and Figure 1C). Although some heterogeneity was observed for this specific variant, several SNPs withinthislocusachievedgenome-widesignificancewithsubstan-tiallylowerheterogeneity(asshowedbyQ  > 0.05,TableS5).ComparableresultswereobtainedwhenanalyzingeachEu-ropeanstudyseparately(FigureS2).rs2984613alsohadsomeeffectinallICH(OR1.21,95%CI1.12–1.38,p ¼ 6.0 3 10  4 ),probably driven by nonlobar cases, and had no effect onlobarICH.NeitheranalysisadjustingforthemostsignificantSNPs (Figure S3B) nor haplotype testing within these lociidentified additional associations of interest. Table 1. Descriptive Characteristics of Participating StudiesCovariateDiscovery ReplicationMulticenter, US European Studies Cincinnati, US Cincinnati, USGOCHA HM-ICH þ VVH-ICH JUHSS LSR GERFHS I & IIERICH/GERFHS III/CCC-GCCCaseCohortControlCohortCaseCohortControlCohortCaseCohortControlCohortCaseCohortControlCohortCaseCohortControlCohortCaseCohortControlCohort Subjects, n 298 457 212 169 122 163 116 153 797 539 1,681 2,261Age, mean (SD) 74 (10) 72 (8) 74 (11) 71 (9) 67 (12) 65 (13) 75 (10) 75 (10) 67 (15) 66 (15) 62 (15) 43 a (26)Female, n (%) 134 (45) 231 (51) 103 (49) 77 (46) 69 (57) 93 (57) 49 (42) 69 (45) 383 (48) 243 (45) 701 (42) 1,113 (49)Hypertension, n (%) 217 (73) 280 (61) 126 (60) 99 (64) 96 (81) 74 (45) 76 (67) 65 (43) 494 (62) 280 (52) 1,009 (60) 738 b (51)Lobar ICH, n (%) 184 (58) – 88 (40) – 51 (39) – 36 (28) – 327 (41) – 484 (29) –Nonlobar ICH, n (%) 132 (42) – 133 (60) – 80 (61) – 94 (72) – 470 (59) – 1,197 (71) –  Abbreviations are as follows: ICH, intracerebral hemorrhage; GOCHA, Genetics of Cerebral Hemorrhage on Anticoagulation Study; HM-ICH, Hospital del Mar Intracerebral Hemorrhage Study; VVH-ICH, Vall d’Hebron Hospital ICH Study; JUHSS, Jagiellonian University Hemorrhagic Stroke Study; LSR, Lund Stroke Register;GERFHS, Genetic and Environmental Risk Factors for Hemorrhagic Stroke Study; ERICH, Ethnic/Racial Variations of Intracerebral Hemorrhage Study; CCC-GCC,Cincinnati Control Cohort - Genomic Control Cohort. a Includes the 819 GCC pediatric control subjects with ages % 18. b Excludes the 819 GCC pediatric control subjects with no hypertension history. 4  The American Journal of Human Genetics  94 , 1–11, April 3, 2014 Please cite this article in press as: Woo et al., Meta-Analysis of Genome-Wide Association Studies Identifies 1q22 as a Susceptibility Locus forIntracerebral Hemorrhage, The American Journal of Human Genetics (2014), http://dx.doi.org/10.1016/j.ajhg.2014.02.012  Replication of identified associations was pursued in1,681 case subjects (513 non-Hispanic whites, 634 AfricanAmericans, and 534 Hispanics) and 2,261 control subjects(1,552 non-Hispanic whites, 449 African Americans, and260 Hispanics) (Tables 1 and S4). Genotyping at this stage included both direct genotyping (ERICH study) andgenome-wide genotyping (GERFHS III and CCC-GCCstudies). Direct genotyping in ERICH included the topSNP at each locus: rs11179580 for 12q21.1 in lobar ICHand rs2758605 for 1q22 in nonlobar ICH. For technicalreasons, rs2758605 (second top SNP) was genotyped inlieu of rs2984613 (r 2 ¼  0.99). The association of chromo-somal region 1q22 with nonlobar ICH replicated but thatof 12q22.1 with lobar ICH did not. No association wasfound between rs11179580, the top SNP at 12q21.1, andlobar ICH (OR 1.05, 95% CI 0.89–1.24; p  ¼  0.55; meta-analysis p ¼ 2.6 3 10  5 ; Figure 2A). For rs2758605 (1q22)in nonlobar ICH, each additional C allele was associated Figure 1. Genome-wide Association Study Results Genome-wide association study results of autosomal SNPs: (A) all (lobar ICH and nonlobar ICH combined), (B) lobar ICH, and (C)nonlobar ICH. The plots show –log10-transformed p values for genotyped and imputed SNPs with respect to their physical positions.The threshold for association at genome-wide significance (p  ¼  5  3  10  8 ) is shown by the upper dashed line, and the lower dashedline corresponds to p ¼ 1 3 10  5 . Landmark genes are indicated for loci that reached the threshold to pursue replication. The American Journal of Human Genetics  94 , 1–11, April 3, 2014  5 Please cite this article in press as: Woo et al., Meta-Analysis of Genome-Wide Association Studies Identifies 1q22 as a Susceptibility Locus forIntracerebral Hemorrhage, The American Journal of Human Genetics (2014), http://dx.doi.org/10.1016/j.ajhg.2014.02.012
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks