Multi-ancestry genome-wide association meta-analysis of Parkinson’s disease

Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 27 na u e gene ics h ps://doi.o g/10.1038/s41588-023-01584-8A icle Mul i-ances y genome-wide associa ion me a-analysis o Pa kinson’s disease Jonggeol Je ey Kim   1,2,167 , Dan Vi ale1,3,4,167, Diego Véliz O ani5,6,167, Michelle Mulan Lian7,8,167, Ka l Heilb on9, he 23andMe Resea ch Team*, Hi o aka Iwaki1,3,4, Julie Lake1, Ca oline Wa ly Solsbe g   10,11,12, Hamp on Leona d1,3,4, Ma y B. Maka ious   1,13,14, Eng-King Tan   15, And ew B. Single on1,4, Sa a Band es-Ciga1,4, Alas ai J. Noyce2, he Global Pa kinson’s Gene ics P og am (GP2)*, Co nelis Blauwend aa    1,4,168 , Mike A. Nalls   1,3,4,168 , Jia Nee Foo7,8,168 & Ignacio Ma a16,168 Al hough o e 90 independen isk a ian s ha e been iden i ied o Pa kinson’s disease using genome-wide associa ion s udies, mos s udies ha e been pe o med in jus one popula ion a a ime. He e we pe o med a la ge-scale mul i-ances y me a-analysis o Pa kinson’s disease wi h 49,049 cases, 18,785 p oxy cases and 2,458,063 con ols including indi iduals o Eu opean, Eas Asian, La in Ame ican and A ican ances y. In a me a-analysis, we iden i ied 78 independen genome-wide signi ican loci, including 12 po en ially no el loci (MTF2, PIK3CA, ADD1, SYBU, IRS2, USP8, PIGL, FASN, MYLK2, USP25, EP300 and PPP6R2) and ine-mapped 6 pu a i e causal a ian s a 6 known PD loci. By combining ou esul s wi h publicly a ailable eQTL da a, we iden i ied 25 pu a i e isk genes in hese no el loci whose exp ession is associa ed wi h PD isk. This wo k lays he g oundwo k o u u e e o s aimed a iden i ying PD loci in non-Eu opean popula ions. Pa kinson’s disease (PD) is a neu odegene a i e disease pa hologi- cally de ined by Lewy body inclusions in he b ain and he dea h o dopamine gic neu ons in he midb ain. The iden i ica ion o gene ic isk ac o s is impe a i e o mi iga ing he global bu den o PD, one o he as es g owing age- ela ed neu odegene a i e diseases. A la ge PD genome-wide associa ion s udy (GWAS) me a-analysis unco e ed 90 independen gene ic isk a ian s in indi iduals o Eu opean ances- y1. Simila ly, la ge-scale PD GWAS me a-analyses o Eas Asian2 and a single GWAS o La in Ame ican 3 indi iduals ha e each iden i ied wo isk loci ha we e no p e iously iden i ied in Eu opeans. Fo PD, he e Recei ed: 26 Augus 2022 Accep ed: 20 Oc obe 2023 Published online: 28 Decembe 2023 Check o upda es 1Labo a o y o Neu ogene ics, Na ional Ins i u e on Aging, Na ional Ins i u es o Heal h, Be hesda, MD, USA. 2P e en i e Neu ology Uni , Cen e o P e en ion Diagnosis and De ec ion, Wol son Ins i u e o Popula ion Heal h, Queen Ma y Uni e si y o London, London, UK. 3Da a Tecnica In e na ional, Washing on, DC, USA. 4Cen e o Alzheime ’s and Rela ed Demen ias (CARD), Na ional Ins i u e on Aging and Na ional Ins i u e o Neu ological Diso de s and S oke, Na ional Ins i u es o Heal h, Be hesda, MD, USA. 5Neu ogene ics Resea ch Cen e , Ins i u o Nacional de Ciencias Neu ológicas, Lima, Pe u. 6Ins i u e o Genome Sciences, Uni e si y o Ma yland, Bal imo e, MD, USA. 7Lee Kong Chian School o Medicine, Nanyang Technological Uni e si y Singapo e, Singapo e, Singapo e. 8Genome Ins i u e o Singapo e, Agency o Science, Technology and Resea ch, A*STAR, Singapo e, Singapo e. 923andMe, Inc., Sunny ale, CA, USA. 10Pha maceu ical Sciences and Pha macogenomics, UCSF, San F ancisco, CA, USA. 11Depa men o Neu ology and Weill Ins i u e o Neu osciences, Uni e si y o Cali o nia, San F ancisco, San F ancisco, CA, USA. 12Memo y and Aging Cen e , UCSF, San F ancisco, CA, USA. 13Depa men o Clinical and Mo emen Neu osciences, UCL Queen Squa e Ins i u e o Neu ology, London, UK. 14UCL Mo emen Diso de s Cen e, Uni e si y College London, London, UK. 15Depa men o Neu ology, Na ional Neu oscience Ins i u e, Duke NUS Medical School, Singapo e, Singapo e. 16Genomic Medicine, Le ne Resea ch Ins i u e, Cle eland Clinic Founda ion, Cle eland, OH, USA. 167These au ho s con ibu ed equally: Jonggeol Je ey Kim, Dan Vi ale, Diego Veliz-O ani, Michelle Mulan Lian. 168These au ho s join ly supe ised his wo k: Co nelis Blauwend aa , Mike A. Nalls, Jia Nee Foo, Ignacio Ma a. *A lis o au ho s and hei a ilia ions appea s a he end o he pape .  e-mail: [email p o ec ed]; co nelis.blauwend aa @nih.go ; mike@da a ecnica.com; jianee. [email p o ec ed].sg; ma ai@cc .o g Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 28 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 PESCA 0.3 ( e . 10) was un o he main Eu opean and Eas Asian me a-analyses and all loci iden i ied in he main analysis we e explo ed (Supplemen a y Table 6). PESCA uses ances y-ma ched LD es ima es o in e whe he he causal a ian s a e popula ion-speci ic o sha ed be ween wo popula ions. Va ian s iden i ied as sha ed be ween he popula ions may be mo e likely o be causal. In addi ion, we expec highe pos e io p obabili y (PP) o sha ed causal a i- an s in he loci iden i ied by MAMA, e en i hey ha e no p e iously been iden i ied in he single-ances y s udy. The lead SNP in he RIMS1 locus ( s12528068) had a high PP o being a sha ed causal a ian (PP = 0.972) despi e being signi ican in he Eu opean s udy1 bu no in he Eas Asian s udy2. We also obse ed ha he no el lead a ian s o MTF2 ( s35940311), PIK3CA ( s11918587), EP300 ( s4820434) and PPP6R2 ( s60708277) had highe PP es ima es o being sha ed causal a ian s ac oss bo h popula ions (PPsha ed = 0.757, 0.214, 0.769, 0.946) han o being causal a ian s in a single popula ion (PP EUR <0.080, PPEAS < 0.001). Howe e , i is impo an o no e ha he sample size disc epancy be ween he Eu opean and Eas Asian da a impac s ou powe o de ec popula ion-speci ic causal a ian s a any o hese loci. We ound 17 sugges i e loci ha ailed o mee ou s ingen sig- ni icance h eshold bu had P < 5 × 10 −8 in a ixed-e ec s me a-analysis and P < 1 × 10−6 in he andom-e ec s me a-analysis (Supplemen a y Table 4). Fou een o hese egions we e no el loci. Two loci nea JAK1 and HS1BP3 we e exclusi ely ound in he 23andMe La in Ame ican and A ican coho s. The lead SNPs ( s578139575 and s73919910) o hese loci a e non-coding and e y a e in Eu opean popula ions bu a e mo e common in A icans and La in Ame icans (gnomAD 3.1.2 mino allele equencies in EUR: 0.02%, 0.23%; AFR: 1.64%, 8.84%; AMR: 0.41%, 1.91%). I con i med, hese loci would con e a s ong e ec on PD isk (be a: −1.3, −0.54). These loci me i u he s udies in he A ican and La in Ame ican popula ions. Fine-mapping iden i ies six c edible se s wi h single a ian s Fine-mapping was also pe o med using MR-MEGA, which uses ances y he e ogenei y o inc ease ine-mapping esolu ion. We iden i ied 23 loci ha had ewe han 5 a ian s wi hin he 95% c edible se . O hese, MR-MEGA nomina ed a single pu a i e causal a ian wi h >95% PP in 6 loci: TMEM163, TMEM175, SNCA, CAMK2D, HIP1R and LSM7 (Table 3 and Supplemen a y Tables 7 and 8). Ou esul s a i med p e ious esul s showing he TMEM175 p.M393T coding a ian as he likely causal a ian 11 . The pu a i e a ian s HIP1R ha e s ong e idence o egulome binding (RegulomeDB ank ≤ 2). In pa icula he HIP1R a ian s10847864 is loca ed in a ansc ip ion s a si e ha is ac i e in subs an- ia nig a issue (ch oma in s a e windows: ch 12:123326200.123327200) and as ocy es in he spinal co d and he b ain (ch oma in s a e windows: ch 12:123326400.123326600). Ou side o he c edible se s con aining a single a ian , we iden i ied missense a ian s in wo genes: FCGR2A (p.H167R, PP = 0.145) and SLC18B1 (p.S30P, PP = 0.780). a e now la ge-scale e o s o sequence and analyze genomic da a in unde ep esen ed popula ions wi h he goal o bo h iden i ying no el associa ed loci, ine-mapping known loci and add essing he inequali y ha exis s in cu en p ecision medicine e o s 4,5 . He e we pe o med a la ge-scale mul i-ances y me a-analysis (MAMA) o PD GWASs by including indi iduals om ou ances al popula- ions: Eu opean, Eas Asian, La in Ame ican and A ican. This e o can se e as a guide o u u e gene ic analyses o inc ease ances al ep esen a ion. Me a-analyses iden i y 66 known and 12 no el loci In addi ion o esul s om p e iously desc ibed Eu opean 1 , Eas Asian 2 and La in Ame ican3 s udies, we also used FinnGen and addi ional da ase s o Eas Asian, La in Ame ican and A ican coho s om 23andMe, Inc (Table 1, Fig. 1 and Supplemen a y Table 1). In o al, we included 49,049 PD cases, 18,618 p oxy cases ( i s -deg ee ela i e wi h PD) and 2,458,063 neu ologically-heal hy con ols. Gene ic co a iance in e cep s om linkage disequilib ium (LD) sco e eg ession6 wi hin ances ies we e close o ze o o nea he 95% con idence in e al, implying ha he e is no sample o e lap be ween he coho s (Supplemen a y Table 1). A e he da a we e ha monized and mapped o genome build hg19, MAMAs we e conduc ed using a andom-e ec s model and me a- eg ession o mul i-e hnic gene ic associa ion (MR-MEGA)7. The andom-e ec s model had g ea e powe o de ec homogenous allelic e ec s 7 . MR-MEGA uses axes o gene ic a ia ion as co a ia es in i s me a- eg ession analysis and had g ea e powe o de ec he e ogeneous e ec s ac oss he di e en coho s. MR-MEGA also dis inguishes ances al he e ogenei y (di e ences in e ec es ima es due o ances y-le el gene ic a ia ion) om esidual he e ogenei y using axes o gene ic a ia ion gene a ed om he allele equencies ac oss he di e en coho s. Combining esul s om he andom-e ec s model and MR-MEGA, we ound 12 no el PD isk loci and 66 hi s in known isk loci om single-ances y GWAS (Table 2, Fig. 2 and Supplemen a y Tables 2–5) ha me he Bon e oni-co ec ed alpha o 5 × 10 −9 , a mo e s ingen h eshold chosen o accoun o he la ge numbe o haplo ypes esul ing om he ances ally di e se da ase s 8 . O he 78 isk loci iden i ied, 69 we e signi ican in he andom-e ec s model, whe eas 3 we e only signi ican in MR-MEGA. Eigh o he no el loci ound by he andom-e ec me hod showed homogeneous e ec s ac oss he ou di e en ances ies. An addi ional no el locus (FASN) iden i ied by he andom-e ec me hod showed homogeneous e ec s in all a ailable popula ions, bu no e ha his a ian ailed quali y con ol in bo h Eas Asian da ase s. The o he h ee loci, iden i ied exclusi ely in MR-MEGA, showed ances ally he e ogeneous e ec s. All h ee loci (IRS2, MYLK2 and USP25) showed e idence o signi ican ances al he e ogenei y (P ANC-HET < 0.05) bu no signi ican esidual he e oge- nei y (PRES-HET > 0.148), suppo ing he idea ha he signals a e due o popula ion s uc u al di e ences a he han o he con ounding ac o s (Fig. 3). Fo he IRS2 locus (lead SNP s1078514, P ANC-HET = 5.3 × 10−3) he Finnish coho has an opposi e e ec di ec ion compa ed o he me a-analysis e ec es ima e (Supplemen a y Fig. 4). Simila ly, he MYLK2 locus has he A ican e ec es ima e mos di e en om he me a-analysis e ec es ima e (lead SNP s6060983, P ANC-HET = 0.035), sugges ing di e en e ec s be ween popula ions. Al hough his is a no el single- ai GWAS locus, i s lead SNP was p e iously disco e ed as a po en ial pleio opic locus in a mul i- ai condi ional/conjunc ional alse disco e y a e (FDR) s udy be ween schizoph enia and PD 9 . Las ly, he USP25 locus had he mos signi ican ances al he e ogenei y (lead SNP s1736020, P ANC-HET = 4.74 × 10 −5 ) and i s e ec s we e speci ic o Eu opean and A ican coho s, albei in di e en di ec ions. When looking a he nea es p o ein coding gene o each no el lead SNP and hei p obabili y o being loss-o - unc ion in ole an (pLI) sco e, we ound ha 7 ou o 12 genes had a pLI sco e o 0.99 o 1. Genes wi h low pLI sco es we e ound bo h in loci wi h (MYLK2) and wi hou (SYBU, PIGL and PPP6R2) signi ican ances y he e ogenei y. Table 1 | Coho desc ip ions S udy Ances al popula ion Cases/p oxy/con ols Nalls e al.1Eu opean (EUR) 37,688/18,618/1,411,006 Foo e al.2Eas Asian (EAS) 6,724/0/24,851 LARGE-PD 3 La in Ame ican (AMR) 807/0/690 FinnGen Release 4 Eu opean-Finnish (EUR) 1,587/0/94,096 23andMe—A ican A ican (AFR) 288/0/193,985 23andMe—Eas Asian Eas Asian (EAS) 322/0/151,905 23andMe—La ino La in Ame ican (AMR) 1,633/0/581,530 MAMA 49,049/18,618/2,458,063 Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 29 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 Gene se analysis inds en ichmen in b ain issues We used he Func ional Mapping and Anno a ion (FUMA) so wa e 12,13 o unc ionally anno a e he andom-e ec esul s. We gene a ed a cus om 1000 Genome e e ence panel ha e lec ed he ances y p opo ions o ou da ase and an mul i-ma ke analysis o genomic anno a ion (MAGMA)14 o gene on ology, issue le el and single-cell exp ession da a. We es ed 16,992 gene on ology se s in MSigDB 7.0 ( e . 15) and used condi ional analysis o disca d edundan e ms o iden i y gene se s ha mus be in e p e ed oge he . We ound ha 40 gene se s we e signi ican ly en iched wi h condi ional analysis iden i ying 13 gene se s ha sha e hei signals wi h a leas one o he gene se (Supplemen a y Table 9). This is a subs an ial inc ease om p e ious 10 gene se s in he Eu opean me a-analysis pe o med by Nalls and colleagues 1 . Only wo gene on ology e ms ha we e signi ican in he Nalls e al. me a-analysis we e also signi ican in he mul i-ances y esul s a e mul iple es co ec ion: ‘cu a ed genese : Ikeda MIR30 Ta ge s Up’ (P FDR = 0.018) and ‘cellula componen : acuola memb ane’ (PFDR = 0.047). In addi ion, on ology e ms in immune sys em pa hways (mic oglial cell p oli e a ion, mac ophage p oli e a ion, na u al kille T cell di e en ia ion: PFDR < 0.04), mi ochond ia ( esponse o mi o- chond ial depola iza ion: PFDR = 0.028), esicles ( esicle uncoa ing, Gene on ology en ichmen Colocaliza ion wi h eQTL 0 2 4 6 8 10 124.2124.1 124.3 124.4 124.5 0.8 0.2 0.4 0.6 2 Fine-mapping Downs eam analyses Goal: In e p e he me a-analysis esul s and iden i y po en ial a ge s and biological mechanisms s11650438 0 100 200 0 2 4 6 8 B ain mmQTL –log10(P) B ain mmQTL –log10(P) PD MAMA –log10(P) PD MAMA –log10(P) s11650438 0 2 4 6 8 0 100 200 15.9 16.0 16.1 16.2 ch 17 (Mb) CENPV 0 5 10 15 –log10(P) 5 10 15 Eu opean 39,275 cases 18,618 p oxy cases 1.5M con ols Eas Asian 7,046 cases 176,756 con ols La ino 2,440 cases 582,220 con ols A ican 288 cases 193,985 con ols S udy pa icipan s Goal: Colla e he la ges and mos di e se se o pa icipan s in Pa kinson’s disease genomics Mul iances y genome-wide me a-analysis Goal: Iden i y common SNPs ha a e associa ed wi h Pa kinson’s disease isk ha a e applicable ac oss di e en ances ies MR-MEGA (he e ogeneous) Random e ec (homogeneous) ++++ +–+– Fig. 1 | MAMA s udy design. Top panel: ou ances y g oups used in he me a- analysis. Middle panel: MAMA and he wo me hods used. Random-e ec ( op) is be e sui ed o isk a ian s wi h homogeneous e ec di ec ion ac oss di e en ances ies, whe eas MR-MEGA (bo om) can iden i y isk a ian s wi h he e ogeneous e ec s due o popula ion s a i ica ion in oduced by ances y di e ences. The densely dashed lines indica e Bon e oni adjus ed sugges i e h eshold o wo-sided P < 1 × 10−6, and he loosely dashed lines indica e Bon e oni adjus ed signi ican h eshold o wo-sided P < 5 × 10−9. Bo om panel: downs eam analyses and hei examples. C ea ed wi h Bio ende .com. Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 30 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 phagolysosome assembly, egula ion o au ophagosome ma u a ion: P FDR < 0.03) and au p o ein ( au p o ein kinase ac i i y: P FDR = 0.034) we e signi ican . A he issue le el, he genes o in e es we e en iched in all b ain cell ypes, as well as pi ui a y issue (Supplemen a y Fig. 9), consis en wi h he esul s om Nalls e al.1. When analyzing single-cell RNA-sequencing da a, he e was no exp ession en ichmen ac oss 88 b ain cell ypes in mouse b ain da a when c oss- e e enced wi h D opViz16 (Supplemen a y Fig. 10). The e was also no en ichmen o any speci ic cell ypes in he subs an ia nig a issue in D opViz (Supplemen a y Fig. 10). Howe e , in human midb ain da a17, dopamine gic (DA1) and GABAe gic (GABA) neu ons we e en iched (Supplemen a y Fig. 10). eQTLs and SMR nomina e 25 pu a i e genes nea no el loci We also sea ched he GTEx 8 ( e . 18) b ain issue eQTLs and mul i-ances y eQTL me a-analysis o he b ain 19 o co ela e no el loci wi h gene exp ession da a (Supplemen a y Tables 10 and 11). To co ela e po en ial pu a i e genes wi h PD isk, we sea ched he signi ican -eQTL genes and genes nea he loci wi h p e iously com- ple ed summa y-based Mendelian andomiza ion (SMR)20 esul s in Eu opean-only da a. When compa ing he SNPs in no el loci wi h mul i-ances y b ain eQTLs19, 28 genes we e signi ican (Supplemen a y Fig. 8 and Supplemen a y Tables 10 and 11). SMR ound 25 genes in ou no el loci associa ed wi h PD isk (Table 2 and Supplemen a y Table 12). In e es ingly, PPP6R2 and CENPV exp ession changes in subs an ia nig a we e associa ed wi h PD isk. PPP6R2 encodes p o ein phospha ase 6 egula o y subuni 2, a egula o y p o ein o p o ein phospha ase 6 ca aly ic subuni (PPP6C), which is in ol ed in he esicle-media ed anspo pa hway. Cen ome e p o ein V (CENPV) is in ol ed in cen- ome e o ma ion and cell di ision. Discussion This s udy is a la ge-scale GWAS me a-analysis o PD ha inco po a es mul iple di e se ances y popula ions. F om he join coho analysis, we iden i ied 66 independen isk loci nea p e iously known PD isk egions and 12 po en ially no el isk loci. O he pu a i e no el loci, nine had homogeneous e ec s and h ee had he e ogeneous e ec s ac oss he di e en coho s. We ound 17 addi ional sugges i e loci using ixed- e ec s me a-analysis h eshold a P < 5 × 10 −8 and andom-e ec s me a- analysis h eshold a P < 1 × 10−6. We ine-mapped 23 loci by le e aging he di e se ances y popula ions. We highligh ed issues and cell ypes associa ed wi h PD isk, which we e consis en wi h p e ious indings 1 . Finally we used SMR o nomina e 25 pu a i e genes nea ou no el loci. No el loci con ained genes in pa hways p e iously implica ed in PD. The MTF2 and PPP6R2 loci con ain he genes TMED5 and PPP6R2. P o ein TMED5 localizes o Golgi body21 and PPP6C, egula ed by PPP6R2, is pa o he esicula anspo pa hways (h ps:// eac ome. o g/con en /de ail/R-HSA-199977)22, bo h o which a e implica ed in PD pa hogenesis23–28. eQTL and SMR analysis showed associa ion be ween exp ession changes o PPP6R2 and CENPV in subs an ia nig a and PD isk. Because subs an ia nig a de e io a ion is a hallma k pa ho- genic ea u e o PD, PPP6R2 and CENPV me i addi ional in es iga ion. Wi hin a known locus, a new independen signal was ound in RILPL2 ( s28659953). P o ein RILPL2 in e ac s wi h LRRK2-phospho yla ed Rab10 o block p ima y cilia gene a ion 29 . Genes JAK1 and HS1BP3 a e in wo sugges i e loci ha we e ound only in La in Ame ican and A ican popula ions. JAK1 is one o he p o eins in he Janus kinase amily, which is a c i ical pa o he JAK-STAT pa hway and is implica ed in cy okine and in lamma o y signaling30. JAK1 a ian s ha e been implica ed in au oimmune diseases such as ju enile idiopa hic a h i is and mul i- ple scle osis31. HS1BP3, also known as essen ial emo 2 (ETM2), has been implica ed in essen ial emo 32–34. Based on i s sequence, ETM2 may modula e in e leukin-2 signaling35. I hese loci a e con i med, hey would u he suppo he g owing app ecia ion o he ole o Table 2 | Me a-analysis esul s o lead SNPs in he no el loci sID Nea es coding gene SMR nomina ed pu a i e genes CHR:BP:A1:A2 BETA(RE) SE P(RE) P(MR-MEGA) P(ANC-HET) P(RES-HET) gnomAD EUR AF gnomAD EAS AF gnomAD AMR AF gnomAD AFR AF pLI s11164870 MTF2 CCDC18 1:93552187:C:G 0.054 0.009 1.15 × 10−10 2.64 × 10−9 0.229 0.928 39.0% 35.1% 45.2% 85.0% 1 s6806917 PIK3CA KCNMB3 3:178861417:T:C −0.070 0.011 1.65 × 10−10 3.43 × 10−9 0.215 0.762 82.0% 89.9% 77.5% 57.8% 1 s16843452 ADD1 ADD1, NOP14-AS1, NOP14 4:2849168:T:C −0.068 0.012 4.11 × 10−9 3.19 × 10−7 0.747 0.687 18.5% 47.4% 18.2% 8.9% 0.99 s6469271 SYBU SYBU 8:110644774:T:C −0.056 0.010 3.62 × 10−9 2.04 × 10−7 0.590 0.954 77.5% 59.3% 74.7% 61.5% 0 s1078514 IRS2 None 13:110463168:T:C 0.068 0.026 4.82 × 10−3 2.30 × 10−95.30 × 10−30.261 33.3% 39.2% 40.6% 10.7% 0.99 s28648524 USP8 TRPM7 15:50787409:A:T 0.064 0.010 6.45 × 10−10 2.58 × 10−8 0.406 0.661 78.1% 53.7% 76.5% 79.8% 1 s11650438 PIGL ADORA2B, ZSWIM7, PIGL, TTC19, NCOR1, CENPV, TRPV2 17:16234260:A:G 0.050 0.009 2.93 × 10−91.46 × 10−7 0.528 0.288 46.9% 17.8% 48.5% 64.0% 0 s4485435 FASN None 17:80045086:C:G 0.082 0.014 2.61 × 10−9N/A N/A N/A 17.3% 12.1% 34.8% 30.3% 1 s6060983 MYLK2 None 20:30420924:T:C 0.069 0.037 0.0322 3.86 × 10−90.035 0.149 69.3% 99.0% 71.8% 29.0% 0.23 s1736020 USP25 None 21:16812552:A:C 0.006 0.005 0.885 1.12 × 10−94.74 × 10−50.638 43.0% 18.6% 38.6% 13.2% 0.75 s73174657 EP300 ZC3H7B, POLR3H, CSDC2, PMM1, RANGAP1, MEI1, L3MBTL2, SLC25A17 22:41434158:A:G −0.059 0.010 3.81 × 10−94.90 × 10−7 0.983 0.655 27.2% 6.3% 47.5% 14.2% 1 s10775809 PPP6R2 PPP6R2 22:50808017:A:T 0.092 0.015 4.09 × 10−10 5.61 × 10−8 0.943 0.903 10.1% 80.3% 80.1% 56.5% 0.16 MR-MEGA could no be un o he lead SNP o he FASN locus, as i was missing in mo e han h ee coho s: Foo e al.2, 23andMe Eas Asian and 23andMe La ino. No P alues we e co ec ed o mul iple es s. CHR, ch omosome; BP, base pai ; A1, e ec allele; A2, o he allele; BETA(RE), allelic e ec in log odds a io; SE, s anda d e o ; P(RE), wo-sided P alue o associa ion om andom e ec ; P(MR-MEGA): wo-sided P alue o associa ion om MR-MEGA (chi-squa ed es wi h d = 4); P(ANC-HET), P alue o he wo-sided ances al he e ogenei y es (chi-squa ed es wi h d = 3); P(RES-HET): P alue o he wo-sided esidual he e ogenei y es (chi-squa ed es wi h d = 3); gnomAD [Ances y] AF, A1 equency epo ed o Eu opeans (EUR), Eas Asians (EAS), Ame indians (AMR) and A icans (AFR) by gnomAD 3.1.2; pLI, p obabili y o being loss-o - unc ion in ole an sco e om gnomAD 2.1.1 o he nea es coding gene (sco e was una ailable o gnomAD 3.1.2); SMR, summa y-based Mendelian andomiza ion; N/A, no a ailable. Bolded a e all signi ican P alues (P < 5 ×10−9 o he wo-sided associa ion es s, P < 0.05 o he he e ogenei y es s). Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 31 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 in lamma ion in PD36. All o he po en ially no el PD loci iden i ied in his analysis will equi e addi ional eplica ion and unc ional alida ion o elucida e hei ole in PD pa hogenesis. P e ious indings in Eu opean popula ions ound ha polygenic isk sco es explained 16–36% o PD he i abili y1. Al hough we did no pe o m simila es s inco po a ing ou no el loci, hey may explain addi ional he i able PD isk. We ound ha 26 o he 66 de ec ed known PD loci had nominally signi ican ances al he e ogenei y (P ANC-HET < 0.05) and 10 emained signi ican a e Bon e oni co ec ion (PANC-HET < 0.05/62 MR-MEGA loci) (Fig. 3 and Supplemen a y Table 3). This he e ogenei y may be caused by di e ences in e ec sizes and allele equencies be ween he di e en popula ions and hus should be s udied as loci wi h po en- ially ances ally di e gen isk. 18 o he p e ious 92 known loci om single-ances y GWASs did no o e lap wi h any genome-wide signi i- can loci in he mul i-ances y esul s a he signi icance h eshold o 5 × 10−9 (Supplemen a y Table 13). Howe e , ou esul s do no neces- sa ily in alida e hese p e ious esul s. Fi s , se e al o he coho s ha e small sample sizes, which may inc ease he in luence o sampling a ia ion. Ano he eason may be due o he s ingen genome-wide signi icance h eshold o 5 × 10 −9 . Al hough his is a la ge PD GWAS me a-analysis, he mo e s ingen signi icance h eshold u he aises he sample size needed o achie e equi alen s a is ical powe . O he 17 Eu opean loci iden i ied, 3 we e signi ican a he 5 × 10−8 h eshold, and all 17 loci we e a leas nominally signi ican wi h he MR-MEGA me hod (P MR-MEGA < 5 × 10 −6 ). Las ly, a ian s may be mo e speci ic o he popula ion in which hey we e i s iden i ied. 5 o he 17 a ian s had nominal ances al he e ogenei y (P ANC-HET < 0.05). I is wo h no ing ha he e a e la ge di e ences in s a is ical powe ac oss ances ies. Addi ional popula ion-speci ic loci will likely each signi icance when la ge sample sizes a e a ailable o non-Eu opean da ase s. Ou ine-mapping isola ed se e al pu a i e causal a ian s in p e iously disco e ed loci. TMEM175- s34311866 has been p e iously iden i ied as unc ionally ele an o PD isk37, which is consis en wi h ou ine-mapping esul s. Fine-mapped a ian s in TMEM163, HIP1R and CAMK3D we e also ound o be pa s o ac i e o s ong ansc ip ion si es in subs an ia nig a issues. Among he ine-mapped a ian s we e wo missense a ian s in FCGR2A and SLC18B1, albei wi h a lowe PP han he 7 singula pu a i e a ian s ha we highligh ed in Table 3. FCGR2A is p esen in mul iple immune- ela ed on ology gene se s, u he highligh ing he po en ial ole o he immune sys em in PD pa hology. Howe e , he unc ion o SLC18B1 is s ill unknown. Al hough he ine-mapping esul s p o ided by MR-MEGA a e su icien a b 0 1 2 MTF2 PIK3CA IRS2 MYLK2 USP25 MTF2 PIK3CA ADD1 SYBU USP8 PIGL FASN EP300 PPP6R2 3 54 6 7 98 Ch omosome 1110 1312 15 16 17 18 19 20 21 2214 1 2 3 54 6 7 98 Ch omosome 1110 1312 15 16 17 18 19 20 21 2214 5 10 15 20 25 –log10(P) 30 35 40 75 0 5 10 15 20 25 –log10(P) 30 35 40 123 90 Fig. 2 | Manha an plo s o he me a-analysis esul s ac oss 2,525,730 pa icipan s. a, Random-e ec s model es . b, MR-MEGA me a- eg ession es (chi-squa ed es wi h d = 4). The x axis shows ch omosome and base pai posi ions o each a ian es ed in he me a-analyses. The y axis shows he wo-sided P alue wi h no mul iple- es co ec ion in he −log10 scale. O ange ho izon al dashed line indica es he Bon e oni-adjus ed signi ican h eshold o P < 5 × 10−9. G ay ho izon al dashed line indica es he unca ion line, whe e all − log10 P alues g ea e han 40 we e unca ed o 40 o isual cla i y. No el loci a e highligh ed in ed and anno a ed wi h he nea es p o ein coding gene. Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 32 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 o iden i y pu a i e causal a ian s o loci d i en by one independen signal, mul iple a ian s in a locus can con ibu e o complex ai s. The addi i e and epis a ic e ec s o mul iple causal a ian s in a locus can be di icul o in e p e when he e ec s associa ed wi h each independen signal a e small. The gene on ology analysis ound mul iple pa hways ha may be ele an o PD pa hology (Supplemen a y Table 9), including hose ela ed o mi ochond ia ( esponse o mi ochond ial depola iza ion) esicles ( esicle uncoa ing, phagolysosome assembly, egula ion o au ophagosome ma u a ion) au p o ein ( au p o ein kinase ac i i y) USP25 ( s1736020-A)* FAM47E ( s6854006-T) ASH1L ( s12734374-A) BRIP1 ( s61169879-T) TMEM163 ( s57891859-A) STK39( s12987123-A) IRS2 ( s1078514-T)* CNTN1 ( s73273590-A) ASXL3 ( s8091977-T) MYLK2 ( s6060983-T)* IGSF9B ( s3802920-T) SV2C ( s246815-C) ITGA2B ( s850731-T) DLG2 ( s6592142-T) ELOVL7 ( s1867598-A) HIP1R ( s10847864-T) SIPA1L2 ( s1326293-T) SETD1A ( s1870293-T) ITGA8 ( s13664-T) IP6K2 ( s12497850-T) PDLIM2 ( s11783129-A) VAMP4 ( s10903-A) MIPOL1 ( s1956438-A) RIMS1 ( s12528068-T) ALL EUR EUR2 EAS EAS2 AMR AMR2 AFR 0 20 40 60 80 100 He e ogenei y (I2, %) I2 Ances y Residual –0.6514 0 0.4003 Be a He e ogenei y in op a ian pe locus (I2 > 30% only) MTF2 ( s11164870-C) PIK3CA ( s7636454-A) ADD1 ( s16843452-T) SYBU ( s6469271-T) IRS2 ( s1078514-T) USP8 ( s28648524-A) PIGL ( s11650438-A) FASN ( s4485435-C) MYLK2 ( s6060983-T) USP25 ( s1736020-A) EP300 ( s73174657-A) PPP6R2 ( s10775809-A) ALL EUR EUR2 EAS EAS2 AMR AMR2 AFR 0 20 40 60 80 100 He e ogenei y (I2, %) I2 Ances y Residual –0.2846 0 0.4003 Be a He e ogenei y in op a ian s in no el loci a b Fig. 3 | He e ogenei y upse plo s. a, Top a ian s pe no el loci. b, Top a ian s pe MR-MEGA iden i ied locus wi h mode a e o high he e ogenei y (I2 > 30). The op ba plo illus a es he e ogenei y wi h da k blue indica ing ances y he e ogenei y p opo ion and ligh blue indica ing o he esidual he e ogenei y p opo ion. The bo om plo shows he subcoho le el be a alues wi h blue indica ing posi i e and ed indica ing nega i e e ec di ec ions. Th ee a ian s wi h g ea e han 30% I2 o al he e ogenei y we e only iden i ied in he MR-MEGA me a-analysis me hod, whe eas li le o no he e ogenei y is obse ed in loci iden i ied in andom e ec . Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 33 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 and immune cells (mic oglial cell/mac ophage p oli e a ion, and na u- al kille T cell di e en ia ion)36. Nei he mi ochond ial no immune cell pa hways we e signi ican in he p e ious Eu opean-only me a-analysis. No el signals om he mul i-ances y app oach may ha e gi en enough powe o highligh hese on ology e ms. Ou o 10 on ology e ms ha we e signi ican in he p e ious Eu opean-only me a-analysis1, 4 e ms we e no es ed due o e sion di e ences in MSigDB and only 2 o he emaining e ms we e signi ican . Howe e , he o he 4 e ms we e s ill nominally signi ican a P < 0.05. This may be due o genome-wide signals ha we e less signi ican due o hei he e ogenei y ac oss he di e en popula ions. Al hough his is a la ge mul i-ances y PD me a-analysis GWAS, he Eu opean popula ion is s ill o e ep esen ed. A ound 80% o ull PD cases a e o Eu opean descen . Indi iduals o A ican descen we e pa icula ly unde ep esen ed a jus 0.5% o he e ec i e PD cases. The disco e ies in ou s udy wa an u u e e o s o expand s udies in mo e di e se popula ions. The Global Pa kinson’s Gene ics P og am (GP2) is pa ne ing wi h ins i u ions ha ca e o unde ep esen ed popula ions o gene a e da a o hese unde se ed communi ies all o e he wo ld5, and we will con inue he ongoing analysis as mo e pa icipan s a e geno yped. Jus as he i s PD GWASs ailed o iden i y signi ican signals38,39, we a e con iden ha u u e di e se ances y GWAS will p oduce impac ul associa ion esul s as sample sizes inc ease. Fu he e o s in mul i-ances y and non-Eu opean GWAS will iden i y loci ha a e mo e ele- an o he global popula ion and will con inue o acili a e ine- mapping e o s o iden i y he gene ic a ian s ha d i e hese associa ions. Online con en Any me hods, addi ional e e ences, Na u e Po olio epo ing sum- ma ies, sou ce da a, ex ended da a, supplemen a y in o ma ion, acknowledgemen s, pee e iew in o ma ion, de ails o au ho con i- bu ions and compe ing in e es s and s a emen s o da a and code a ail- abili y a e a ailable a h ps://doi.o g/10.1038/s41588-023-01584-8. Re e ences 1. Nalls, M. A. e al. Iden i ica ion o no el isk loci, causal insigh s, and he i able isk o Pa kinson’s disease: a me a-analysis o genome-wide associa ion s udies. Lance Neu ol. 18, 1091–1102 (2019). 2. Foo, J. N. e al. Iden i ica ion o isk loci o Pa kinson disease in Asians and compa ison o isk be ween Asians and Eu opeans: a genome-wide associa ion s udy. JAMA Neu ol. 77, 746–754 (2020). 3. Loesch, D. P. e al. Cha ac e izing he gene ic a chi ec u e o Pa kinson’s disease in La inos. Ann. Neu ol. 90, 353–365 (2021). 4. Rajan, R. e al. Gene ic a chi ec u e o Pa kinson’s disease in he Indian popula ion: ha nessing gene ic di e si y o add ess c i ical gaps in Pa kinson’s disease esea ch. F on . Neu ol. 11, 524 (2020). 5. Global Pa kinson’s Gene ics P og am. GP2: The Global Pa kinson’s Gene ics P og am. Mo . Diso d. 36, 842–851 (2021). 6. Bulik-Sulli an, B. K. e al. LD Sco e eg ession dis inguishes con ounding om polygenici y in genome-wide associa ion s udies. Na . Gene . 47, 291–295 (2015). 7. Mägi, R. e al. T ans-e hnic me a- eg ession o genome-wide associa ion s udies accoun ing o ances y inc eases powe o disco e y and imp o es ine-mapping esolu ion. Hum. Mol. Gene . 26, 3639–3650 (2017). 8. Puli , S. L., de Wi h, S. A. J. & de Bakke , P. I. W. Rese ing he ba : s a is ical signi icance in whole-genome sequencing-based associa ion s udies o global popula ions. Gene . Epidemiol. 41, 145–151 (2017). 9. Smeland, O. B. e al. Genome-wide associa ion analysis o Pa kinson’s disease and schizoph enia e eals sha ed gene ic a chi ec u e and iden i ies no el isk loci. Biol. Psychia y 89, 227–235 (2021). 10. Shi, H. e al. Localizing componen s o sha ed anse hnic gene ic a chi ec u e o complex ai s om GWAS summa y da a. Am. J. Hum. Gene . 106, 805–817 (2020). 11. Jinn, S. e al. Func ionaliza ion o he TMEM175 p.M393T a ian as a isk ac o o Pa kinson disease. Hum. Mol. Gene . 28, 3244–3254 (2019). 12. Wa anabe, K., Taskesen, E., an Bocho en, A. & Pos huma, D. Func ional mapping and anno a ion o gene ic associa ions wi h FUMA. Na . Commun. 8, 1826 (2017). 13. Wa anabe, K., Umiće ić Mi ko , M., de Leeuw, C. A., an den Heu el, M. P. & Pos huma, D. Gene ic mapping o cell ype speci ici y o complex ai s. Na . Commun. 10, 3222 (2019). 14. de Leeuw, C. A., Mooij, J. M., Heskes, T. & Pos huma, D. MAGMA: gene alized gene-se analysis o GWAS da a. PLoS Compu . Biol. 11, e1004219 (2015). 15. Libe zon, A. e al. Molecula signa u es da abase (MSigDB) 3.0. Bioin o ma ics 27, 1739–1740 (2011). 16. Saunde s, A. e al. Molecula di e si y and specializa ions among he cells o he adul mouse b ain. Cell 174, 1015–1030.e16 (2018). 17. La Manno, G. e al. Molecula di e si y o midb ain de elopmen in mouse, human, and s em cells. Cell 167, 566–580.e19 (2016). 18. GTEx Conso ium. The GTEx Conso ium a las o gene ic egula o y e ec s ac oss human issues. Science 369, 1318–1330 (2020). 19. Zeng, B. e al. Mul i-ances y eQTL me a-analysis o human b ain iden i ies candida e causal a ian s o b ain- ela ed ai s. Na . Gene . 54, 161–169 (2022). 20. Wu, Y. e al. In eg a i e analysis o omics summa y da a e eals pu a i e mechanisms unde lying complex ai s. Na . Commun. 9, 918 (2018). 21. Koegle , E. e al. p28, a no el ERGIC/cis Golgi p o ein, equi ed o Golgi ibbon o ma ion. T a ic 11, 70–89 (2010). Table 3 | MR-MEGA ine-mapping esul s o loci wi h a single SNP wi hin he 95% c edible se Locus Numbe o signi ican SNPs Nomina ed a ian CHR:BP:A1:A2 Nea es gene Known PD gene ± 1 MB Func ional consequence CADD RDB 11 6 s57891859 2:135464616:A:G TMEM163 TMEM163 in onic 6.746 4 19 926 s34311866 4:951947:C:T TMEM175 TMEM175 exonic 11.09 NA 23 1483 s356182 4:90626111:A:G SNCA SNCA ncRNA in onic 8.962 NA 24 121 s13117519 4:114369065:T:C CAMK2D CAMK2D in e genic 1.216 3a 45 1371 s10847864 12:123326598:G:T HIP1R HIP1R in onic 2.403 2b 60 1 s55818311 19:2341047:C:T SPPL2B LSM7 ncRNA exonic 1.096 5 Known PD genes a e ei he known PD isk genes (SNCA and TMEM175) o genes wi h he highes sco e in he nea es known PD locus by he PD GWAS Locus B owse 37. CHR, ch omosome; BP, base pai ; A1, e ec allele; A2, o he allele; CADD, combined anno a ion-dependen deple ion sco e; RDB, egulomeDB sco e; ncRNA, non-coding RNA. Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 34 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 22. Gaude , P., Li s one, M. S., Lewis, S. E. & Thomas, P. D. Phylogene ic- based p opaga ion o unc ional anno a ions wi hin he Gene On ology conso ium. B ie . Bioin o m. 12, 449–462 (2011). 23. Gillespie, M. ER o Golgi An e og ade T anspo . Reac ome, Release 73 (Reac ome, accessed 29 Ma ch 2021); h ps:// eac ome.o g/con en /de ail/R-HSA-199977 24. Bone -Ponce, L. e al. LRRK2 media es ubula ion and esicle so ing om lysosomes. Sci. Ad . 6, eabb2454 (2020). 25. Band es-Ciga, S. e al. The endocy ic memb ane a icking pa hway plays a majo ole in he isk o Pa kinson disease. Mo . Diso d. 34, 460–468 (2019). 26. Beilina, A. e al. Unbiased sc een o in e ac o s o leucine- ich epea kinase 2 suppo s a common pa hway o spo adic and amilial Pa kinson disease. P oc. Na l Acad. Sci. USA. 111, 2626–2631 (2014). 27. MacLeod, D. A. e al. RAB7L1 in e ac s wi h LRRK2 o modi y in aneu onal p o ein so ing and Pa kinson’s disease isk. Neu on 77, 425–439 (2013). 28. Fuji a, Y., Ohama, E., Taka ama, M., Al-Sa aj, S. & Okamo o, K. F agmen a ion o Golgi appa a us o nig al neu ons wi h alpha-synuclein-posi i e inclusions in pa ien s wi h Pa kinson’s disease. Ac a Neu opa hol. 112, 261–265 (2006). 29. Ma ínez-Mená guez, J. Á., Tomás, M., Ma ínez-Ma ínez, N. & Ma ínez-Alonso, E. Golgi agmen a ion in neu odegene a i e diseases: is he e a common cause? Cells 8, 748 (2019). 30. Sobu, Y., Waw o, P. S., Dhekne, H. S., Yeshaw, W. M. & P e e , S. R. Pa hogenic LRRK2 egula es cilia ion p obabili y ups eam o au ubulin kinase 2 ia Rab10 and RILPL1 p o eins. P oc. Na l Acad. Sci. USA. 118, e2005894118 (2021). 31. Philips, R. L. e al. The JAK-STAT pa hway a 30: much lea ned, much mo e o do. Cell 185, 3857–3876 (2022). 32. Deng, H. e al. Ex ended s udy o A265G a ian o HS1BP3 in essen ial emo and Pa kinson disease. Neu ology 65, 651–652 (2005). 33. Higgins, J. J. e al. A a ian in he HS1-BP3 gene is associa ed wi h amilial essen ial emo . Neu ology 64, 417–421 (2005). 34. Siokas, V. e al. Gene ic isk ac o s o essen ial emo : a e iew. T emo O he Hype kine . Mo . 10, 4 (2020). 35. The UniP o Conso ium e al. UniP o : he uni e sal p o ein knowledgebase in 2021. Nucleic Acids Res. 49, D480–D489 (2020). 36. Kung, P. J., Elsayed, I., Reyes-Pé ez, P. & Band es-Ciga, S. Immunogene ic de e minan s o Pa kinson’s disease e iology. J. Pa kinsons. Dis. 12, S13–S27 (2022). 37. G enn, F. P. e al. The Pa kinson’s disease genome-wide associa ion s udy locus b owse . Mo . Diso d. 35, 2056–2067 (2020). 38. Fung, H.-C. e al. Genome-wide geno yping in Pa kinson’s disease and neu ologically no mal con ols: i s s age analysis and public elease o da a. Lance Neu ol. 5, 911–916 (2006). 39. Ma agano e, D. M. e al. High- esolu ion whole-genome associa ion s udy o Pa kinson disease. Am. J. Hum. Gene . 77, 685–693 (2005). Publishe ’s no e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a ilia ions. Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons license, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle’s C ea i e Commons license and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew a copy o his license, isi h p://c ea i ecommons.o g/licenses/by/4.0/. This is a U.S. Go e nmen wo k and no unde copy igh p o ec ion in he US; o eign copy igh p o ec ion may apply 2023 he 23andMe Resea ch Team S ella Aslibekyan9, Adam Au on9, Elizabe h Babalola9, Robe K. Bell9, Jessica Bielenbe g9, Ka a zyna B yc9, Emily Bullis9, Paul Cannon9, Daniella Coke 9, Gab iel Cuella Pa ida9, De ika Dhamija9, Sayan an Das9, Sa ah L. Elson9, Nicholas E iksson9, Te esa Filsh ein9, Alison Fi ch9, Kippe Fle ez-B an 9, Pie e Fon anillas9, Will F eyman9, Julie M. G anka9, Ka l Heilb on9, Alejand o He nandez9, Ba y Hicks9, Da id A. Hinds9, E han M. Jewe 9, Yunxuan Jiang9, Ka elyn Kuka 9, Alan Kwong9, Keng-Han Lin9, Bianca A. Llamas9, Maya Lowe9, Jey C. McC eigh 9, Ma hew H. McIn y e9, S e en J. Michele i9, Meghan E. Mo eno9, P iyanka Nandakuma 9, Dominique T. Nguyen9, Elizabe h S. Noblin9, Ja ed O’Connell9, Aa on A. Pe ako i z9, G. Da id Poznik9, Alexand a Reynoso9, Madeleine Schloe e 9, Mo gan Schumache 9, Anjali J. Shas i9, Janie F. Shel on9, Jingchunzi Shi9, Suyash Sh inga pu e9, Qiaojuan Jane Su9, Susana A. Ta 9, Ch is ophe Toukam Tchakou é9, Vinh T an9, Joyce Y. Tung9, Xin Wang9, Wei Wang9, Ca he ine H. Weldon9, Pe e Wil on9 & Co inna D. Wong9 he Global Pa kinson’s Gene ics P og am (GP2) Emilia M. Ga o17, Ma celo Kau man18, Samson Khacha yan19, Za uhi Ta adyan19, Clai e E. Shephe d20, Julie Hun e 21, Kisho e Kuma 22, Melina Ellis23, Miguel E. Ren e ía24, Sule Koks25, Alexande Zimp ich26, A u F. Schumache -Schuh27, Ca los Riede 28, Paula Sa ie Awad29, Vi o Tumas30, Sa ah Cama gos31, Edwa d A. Fon32, Ou y Monchi33, Ted Fon34, Benjamin Piza o Galleguillos35, Ma celo Mi anda36, Ma ia Leono Bus aman e37, Pa icio Olguin35, Ped o Chana38, Beisha Tang39, Hui ang Shang40, Ji eng Guo41, Piu Chan42, Wei Luo43, Gonzalo A boleda44, Jo ge O ozc45, Ma lene Jimenez del Rio46, Al a o He nandez47, Mohamed Salama48, Walaa A. Kamel49, Ya ed Z. Zewde50, Alexis B ice51, Jean-Ch is ophe Co ol52, Ana Wes enbe ge 53, Anas asia Illa iono a54, B i Mollenhaue 55, Ch is ine Klein53, E a-Juliane Volls ed 53, F anziska Hop ne 56, Gün e Höglinge 56, Ha u yun Madoe 53, Joanne T inh53, Johanna Junke 53, Ka ja Lohmann53, La a M. Lange53,57, Manu Sha ma58, Se giu G oppa59, Thomas Gasse 58, Zih-Hua Fang60, Albe Akpalu61, Geo gia Xi ome isiou62, Geo gios Hadjigo giou62, Ioannis Dagklis63, Ioannis Ta nanas64, Leonidas S e anis65, Ma ia S amelou66, E hymios Dadio is62, Alex Medina67, Ge maine Hiu-Fai Chan68, Nancy Ip69, Nelson Yuk-Fai Cheung68, Phillip Chan69, Xiaopu Zhou69, Asha Kisho e70, K. P. Di ya71, P amod Pal72, P ashan h Lingappa Kukkle73, Roopa Rajan74, Rupam Bo gohain75, Meh i Sala i76, And ea Qua one77, Na u e Gene ics | Volume 56 | Janua y 2024 | 27–36 35 A icle h ps://doi.o g/10.1038/s41588-023-01584-8 Enza Ma ia Valen e78, Lucilla Pa ne i79, Micol A enali78, Tommaso Schi inzi80, Manabu Funayama81, Nobu aka Ha o i82, Tomo aka Shi aishi83, Al ynay Ka imo a84, Gulnaz Kaishibaye a84, Cholpon Shambe o a85, Rejko K üge 86, Ai Huey Tan87, Azlina Ahmad-Annua 87, Mohamed Ib ahim No linah88, No Azian Abdul Mu ad89, Shah ul Azmin90, Shen-Yang Lim87, Wael Mohamed91, Yi Wen Tay87, Daniel Ma inez-Rami ez92, Mayela Rod iguez-Violan e93, Paula Reyes-Pé ez94, Bayasgalan Tse ensodnom95, Rajee Ojha96, Tim J. Ande son97, Toni L. Pi che 97, A inola Sanyaolu98, Njideka Okubadejo98, Oluwadamilola Ojo99, Jan O. Aasly100, Lasse Pihls øm101, Manuela Tan101, Shoaib U -Rehman102, Diego Veliz-O ani5,6, Ma io Co nejo-Oli as103, Ma ia Leila Doquenia104, Raymond Rosales104, Angel Vinuela105, Elena Iako enko106, Bashaye Al Muba ak107, Muhammad Umai 108, Eng-King Tan15, Michelle Mulan Lian7,8,167, Jia Nee Foo7,8,168, Fe zana Amod109, Jona han Ca 110, So aya Ba dien110, Beomseok Jeon111, Yun Joong Kim112, Es he Cubo113, Ignacio Al a ez114, Jane Hoenicka115, Ka in Beye 116, Ma ia Te esa Pe iñan117, Pau Pas o 118, Sa ah El-Sadig119, Kajsa B olin120, Ch is iane Zweie 121, Ge d Tinkhause 122, Paul K ack121, Chin-Hsien Lin123, Hsiu-Chuan Wu124, Pin-Jui Kung125, Ruey-Meei Wu123, Yih u Wu124, Rim Amou i126, Samia Ben Sassi127, A. Nazl Başak128, Gence Genc129, Özgü Öz op Çakmak128, Sibel E an128, Alas ai J. Noyce2, Alejand o Ma ínez-Ca asco13, Ane e Sch ag13, An hony Schapi a13, Camille Ca oll130, Clai e Bale131, Donald G osse 132, Eleano J. S a o d13, Hen y Houlden13, Huw R. Mo is13,14, John Ha dy13, Kin Ying Mok13, Mie Rizig13, Nicholas Wood13, Nigel Williams133, Olai an Okunoye13, Pa ick Al yn Lewis134, Rauan Kaiy zhano 13, Rimona Weil13, Se h Lo e135, Simon S o 136, Simona Jasai y e13, Sumi Dey2, Vida Obese13, Albe o Espay137, Alyssa O’G ady138, And ew B. Single on1,4, And ew K. Sobe ing139, Be nade e Siddiqi138, B ad o d Casey138, B ian Fiske138, Cabell Jonas140, Ca los C uchaga141, Ca oline B. Pan azis4, Cha isse Coma 138, Clai e Wegel142, Co nelis Blauwend aa 1,4,168, Dan Vi ale1,3,4,167, Debo ah Hall143, Dena He nandez1, Ejaz Shiamim144, Ekemini Riley145, Fa az Fagh i3,4, Geidy E. Se ano146, Hamp on Leona d1,3,4, Hi o aka Iwaki1,3,4, Honglei Chen147, Ignacio F. Ma a16, Ignacio Juan Kelle Sa mien o148, Ja ed Williamson144, Jonggeol Je ey Kim1,2,167, Joseph Janko ic149, Joshua Shulman149,150, Jus in C. Solle138, Kaileigh Mu phy138, Ka en Nuy emans151, Ka l Kiebu z152, Ka e ina Ma kopoulou153, Kenne h Ma ek154, K is in S. Le ine3,4, Lana M. Chahine155, Lau a Ibanez156, Lau el Sc e en4, Lau en Ru age157, Lisa Shulman158, Luca Ma sili137, Maggie Kuhl138, Ma issa Dean157, Ma y B. Maka ious1,13,14, Ma hew Ko e sky1,4, Megan J. Puckelwa z148, Miguel Inca-Ma inez16, Mike A. Nalls1,3,4,168, Naomi Louie138, Niccolò Emanuele Mencacci148, Roge Albin159, Roy Alcalay160, Ru h Walke 161, Sa a Band es-Ciga1,4, Sohini Chowdhu y138, Sonya Dumanis162, S e en Lubbe148, Tao Xie163, Ta iana Fo oud164, Thomas Beach146, Todd She e 138, Yeajin Song3,4, Duan Nguyen165, Toan Nguyen165 & Masha ip A adzhano 166 17Sana o io de la T inidad Mi e- INEBA, Buenos Ai es, A gen ina. 18Hospi al JM Ramos Mejia, Buenos Ai es, A gen ina. 19Somnus Neu ology Clinic, Ye e an, A menia. 20Neu oscience Resea ch Aus alia, Sydney, New Sou h Wales, Aus alia. 21ANZAC Resea ch Ins i u e, Conco d, New Sou h Wales, Aus alia. 22Ga an Ins i u e o Medical Resea ch and Conco d Repa ia ion Gene al Hospi al, Da linghu s , New Sou h Wales, Aus alia. 23Conco d Hospi al, Conco d, New Sou h Wales, Aus alia. 24QIMR Be gho e Medical Resea ch Ins i u e, He s on, Queensland, Aus alia. 25Mu doch Uni e si y, Pe h, Wes e n Aus alia, Aus alia. 26Medical Uni e si y Vienna Aus ia, Vienna, Aus ia. 27Uni e sidade Fede al do Rio G ande do Sul / Hospi al de Clínicas de Po o Aleg e, Po o Aleg e, B azil. 28Fede al Uni e si y o Heal h Sciences o Po o Aleg e, Po o Aleg e, B azil. 29Uni e sidade Fede al do Rio G ande do Sul, Po o Aleg e, B azil. 30Uni e si y o São Paulo, São Paulo, B azil. 31Uni e sidade Fede al de Minas Ge ais, Belo Ho izon e, B azil. 32Mon eal Neu ological Ins i u e, Mon eal, Quebec, Canada. 33Ins i u uni e si ai e de gé ia ie de Mon éal, Mon eal, Quebec, Canada. 34McGill Uni e si y, Mon eal, Quebec, Canada. 35Uni e sidad de Chile, San iago, Chile. 36Fundación Diagnosis, San iago, Chile. 37Facul y o Medicine Uni e sidad de Chile, San iago, Chile. 38CETRAM, San iago, Chile. 39Cen al Sou h Uni e si y, Changsha, China. 40Wes China Hospi al Sichuan Uni e si y, Chengdu, China. 41Xiangya Hospi al, Changsha, China. 42Capi al Medical Uni e si y, Beijing, China. 43Zhejiang Uni e si y, Hangzhou, China. 44Uni e sidad Nacional de Colombia, Bogo á, Colombia. 45Fundación Valle del Lili, San iago De Cali, Colombia. 46Uni e si y o An ioquia, Medellin, Colombia. 47Uni e si y o Cos a Rica, San Jose, Cos a Rica. 48The Ame ican Uni e si y in Cai o, Cai o, Egyp . 49Beni-Sue Uni e si y, Beni Sue , Egyp . 50Addis Ababa Uni e si y, Addis Ababa, E hiopia. 51Pa is B ain Ins i u e, Pa is, F ance. 52So bonne Uni e si é, Pa is, F ance. 53Uni e si y o Lübeck, Lübeck, Ge many. 54Deu sches Zen um ü Neu odegene a i e E k ankungen, Gö ingen, Ge many. 55Uni e si y Medical Cen e Gö ingen, Gö ingen, Ge many. 56Depa men o Neu ology, Uni e si y Hospi al, LMU Munich, Munich, Ge many. 57Uni e si y Medical Cen e Schleswig-Hols ein, Lübeck, Ge many. 58Uni e si y o Tubingen, Tübingen, Ge many. 59Uni e si y o Mainz, Mainz, Ge many. 60The Ge man Cen e o Neu odegene a i e Diseases, Gö ingen, Ge many. 61Uni e si y o Ghana Medical School, Acc a, Ghana. 62Uni e si y o Thessaly, Volos, G eece. 63A is o le Uni e si y o Thessaloniki, Thessaloniki, G eece. 64Ionian Uni e si y, Co u, G eece. 65Biomedical esea ch Founda ion o he Academy o A hens, A hens, G eece. 66Diagnos ic and The apeu ic Cen e HYGEIA Hospi al, Ma ousi, G eece. 67Hospi al San Felipe, Tegucigalpa, Hondu as. 68Queen Elizabe h Hospi al, Kowloon, Hong Kong. 69The Hong Kong Uni e si y o Science and Technology, Kowloon, Hong Kong. 70As e Medci y, Kochi, India. 71S ee Chi a Ti unal Ins i u e o Medical Sciences and Technology, Thi u anan hapu am, India. 72Na ional Ins i u e o Men al Heal h & Neu osciences, Bengalu u, India. 73Manipal Hospi al, Delhi, India. 74All India Ins i u e o Medical Sciences, Delhi, India. 75Nizam’s Ins i u e o Medical Sciences, Hyde abad, India. 76Shahid Behesh i Uni e si y o Medical Science, Teh an, I an. 77Magna G æcia Uni e si y o Ca anza o, Ca anza o, I aly. 78Uni e si y o Pa ia, Pa ia, I aly. 79Uni e si y o Pe ugia, Pe ugia, I aly. 80Uni e si y o Rome To Ve ga a, Rome, I aly. 81Jun endo Uni e si y, Tokyo, Japan. 82Facul y o Medicine, Jun endo Uni e si y, Tokyo, Japan. 83Jikei Uni e si y School o Medicine, Tokyo, Japan. 84Ins i u e o Neu ology and Neu o ehabili a ion, Alma y, Kazakhs an. 85Ky gyz S a e Medical Academy, Bishkek, Ky gyzs an. 86Uni e si y o Luxembou g, Luxembou g, Luxembou g. 87Uni e si y o Malaya, Kuala Lumpu , Malaysia. 88Uni e si i Kebangsaan Malaysia, Selango , Malaysia. 89UKM Medical Molecula Biology Ins i u e, Kuala Lumpu , Malaysia. 90Uni e si i Kebangsaan Malaysia Medical Cen e, Kuala Lumpu , Malaysia. 91In e na ional Islamic Uni e si y, Kuala Lumpu , Malaysia. 92Tecnologico de Mon e ey, Mon e ey, Mexico. 93Ins i u o Nacional de Neu ologia y Neu oci ugia, Mexico Ci y, Mexico. 94Uni e sidad Nacional Au ónoma de México, Mexico Ci y, Mexico. 95Mongolian Na ional Uni e si y o Medical Sciences, Ulaanbaa a , Mongolia. 96T ibhu an Uni e si y, Ki ipu , Nepal. 97Uni e si y o O ago, Dunedin, New Zealand. 98Uni e si y o Lagos, Lagos, Nige ia. 99College o Medicine o he Uni e si y o Lagos, Lagos, Nige ia. 100No wegian Uni e si y o Science and Technology, T ondheim, No way. 101Oslo Uni e si y Hospi al, Oslo, No way. 102Uni e si y o Science and Technology Bannu, Bannu, Pakis an. 103Uni e sidad Cien i ica del Su , Lima, Pe u. 104Me opoli an Medical Cen e , Manila, Philippines. 105Uni e si y o Pue o Rico, San Juan, Pue o Rico. 106Resea ch Cen e o Neu ology, Moscow, Russia. 107King Faisal Specialis Hospi al and Resea ch Cen e , Riyadh, Saudi A abia. 108King Abdullah In e na ional Medical Resea ch Cen e , Jeddah, Saudi A abia. 109Uni e si y o KwaZulu-Na al, Du ban, Sou h A ica. 110S ellenbosch Uni e si y, S ellenbosch, Sou h A ica. 111Seoul Na ional Uni e si y Hospi al, Seoul, Sou h Ko ea. 112Yongin Se e ance Hospi al, Seoul, Sou h Ko ea. 113Hospi al Uni e si a io Bu gos, Bu gos, Spain. 114Uni e si y Hospi al Mu ua Te assa, Ba celona, Spain.