Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms

Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms View ORCID Profile Chen-Yang Su , Juliano Malizia , View ORCID Profile Masashi Hasebe , View ORCID Profile Thomas Zheng , View ORCID Profile Alejandro-Mejia Garcia , View ORCID Profile Hsuan Megan Tsao , Zhaohe Lin , Ta-Yu Yang , View ORCID Profile Fumihiko Matsuda , View ORCID Profile Patrick A. Dion , View ORCID Profile Vincent Mooser , View ORCID Profile Guy Rouleau , View ORCID Profile Guillaume Butler-Laporte , View ORCID Profile Tianyuan Lu , View ORCID Profile Satoshi Yoshiji , View ORCID Profile Sirui Zhou doi: https://doi.org/10.1101/2025.11.11.25339992 Chen-Yang Su 1 Quantitative Life Sciences, McGill University , Montréal, Québec, Canada 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Chen-Yang Su Juliano Malizia 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Masashi Hasebe 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada 5 Department of Diabetes , Endocrinology and Nutrition, Kyoto University Graduate School of Medicine , Kyoto, Japan Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Masashi Hasebe Thomas Zheng 1 Quantitative Life Sciences, McGill University , Montréal, Québec, Canada 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thomas Zheng Alejandro-Mejia Garcia 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Alejandro-Mejia Garcia Hsuan Megan Tsao 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Hsuan Megan Tsao Zhaohe Lin 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ta-Yu Yang 6 Center for Genomic Medicine, Graduate School of Medicine, Kyoto University , Kyoto, Japan Find this author on Google Scholar Find this author on PubMed Search for this author on this site Fumihiko Matsuda 6 Center for Genomic Medicine, Graduate School of Medicine, Kyoto University , Kyoto, Japan Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Fumihiko Matsuda Patrick A. Dion 7 Montreal Neurological Institute-Hospital, McGill University , Montréal, Québec, Canada 8 Department of Neurology and Neurosurgery, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Patrick A. Dion Vincent Mooser 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vincent Mooser Guy Rouleau 7 Montreal Neurological Institute-Hospital, McGill University , Montréal, Québec, Canada 8 Department of Neurology and Neurosurgery, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guy Rouleau Guillaume Butler-Laporte 1 Quantitative Life Sciences, McGill University , Montréal, Québec, Canada 9 Lady Davis Institute, Jewish General Hospital, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Guillaume Butler-Laporte Tianyuan Lu 10 Department of Population Health Sciences, University of Wisconsin-Madison , Madison, WI, USA 11 Department of Biostatistics and Medical Informatics, University of Wisconsin-Madison , Madison, WI, USA 12 Center for Genomic Science Innovation, University of Wisconsin-Madison , Madison, WI, USA 13 Center for Demography of Health and Aging, University of Wisconsin-Madison , Madison, WI, USA 14 Center for Human Genomics and Precision Medicine, University of Wisconsin-Madison , Madison, WI, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Tianyuan Lu Satoshi Yoshiji 1 Quantitative Life Sciences, McGill University , Montréal, Québec, Canada 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada 9 Lady Davis Institute, Jewish General Hospital, McGill University , Montréal, Québec, Canada 15 Programs in Metabolism and Medical & Population Genetics, The Broad Institute of MIT and Harvard , Cambridge, MA, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Satoshi Yoshiji Sirui Zhou 1 Quantitative Life Sciences, McGill University , Montréal, Québec, Canada 2 McGill Genome Centre, McGill University , Montréal, Québec, Canada 3 Canada Excellence Research Chair in Genomic Medicine, McGill University , Montréal, Québec, Canada 4 Department of Human Genetics, McGill University , Montréal, Québec, Canada Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sirui Zhou For correspondence: sirui.zhou{at}mcgill.ca Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Background Intracranial aneurysm (IA) and its complication, subarachnoid hemorrhage (SAH), cause morbidity and mortality, yet no preventative pharmacotherapies exist. Genome-wide association studies (GWAS) have identified risk loci for IA and SAH, but the causal proteins and pathways that connect genetic risk to aneurysm biology remain unclear. Methods We performed ancestry-stratified GWAS meta-analyses of IA and SAH in European and East Asian ancestries and linked these to circulating protein levels using proteome-wide Mendelian randomization (MR). We applied stringent instrument selection, sensitivity analyses, and colocalization, and implemented GWAS-by-subtraction to derive IA components not fully mediated by systolic blood pressure (SBP). We triangulated findings with UK Biobank observational associations, rare variant gene-burden testing in 426,295 exomes, and a French-Canadian familial IA cohort. Results Across 15,611 protein-outcome tests, 12 associations for nine proteins were significant in European ancestry. SLMAP, AMBP, ENTPD6, and PLEKHA1 were associated with increased IA risk, whereas SIRT2, JAG1, ADH4, and NAGLU were associated with decreased IA risk; ADAM23 was associated with increased SAH risk. Colocalization supported shared causal variants for ADH4, PLEKHA1, and SLMAP in IA. After removing SBP-mediated genetic effects, ADH4, JAG1, and PLEKHA1 remained associated with IA, suggesting effects not fully mediated by blood pressure. In UK Biobank, higher measured SLMAP, AMBP, and ENTPD6 levels showed concordant increases in cerebrovascular disease risk. Rare damaging JAG1 variants showed nominally higher odds of cerebrovascular disease, and a missense ENTPD6 variant was enriched in French-Canadian familial IA. Conclusions Integrating multi-ancestry genomics with large-scale proteomics implicates specific circulating proteins and pathways in IA and SAH risk. Convergent evidence prioritizes ADH4, a retinoid-pathway enzyme, and PLEKHA1, a phosphoinositide-binding adaptor in endothelial signaling, as non-SBP-mediated candidates for IA biology, with additional support for JAG1/Notch and SLMAP-related vascular pathways. These findings highlight mechanistic biomarkers and potential drug targets for aneurysm prevention that warrant experimental validation. Introduction Intracranial aneurysms (IA) are cerebrovascular abnormalities caused by localized dilations of cerebral arteries due to vessel wall weakness or thinning 1 . Most intracranial aneurysms are asymptomatic at the time of detection, with a prevalence of 3% in the general population 2 , and lead to an elevated risk of subarachnoid hemorrhage (SAH) resulting from their rupture. SAH has severe consequences and is often fatal, with higher incidence reported in parts of Europe and East Asia 3 , 4 . Among survivors, long-term neurological morbidity is common and results in a substantial burden on healthcare systems. Despite the clinical significance of IA and SAH, there are no approved pharmacological therapies, other than targeting clinical risk factors, to halt aneurysm formation, growth, or rupture. Current curative treatment options are restricted to invasive methods 5 , specifically endovascular procedures, which involve a risk of procedure-related complications 6 . Thus, elucidating the biological mechanisms underlying aneurysm formation and rupture is critical to identify biomarkers therapeutic targets that could reduce the long-term burden of IA. Genome-wide association studies (GWAS) have identified susceptibility loci associated with IA and aneurysmal SAH, highlighting a polygenic contribution to disease risk. Notably, a recent large GWAS meta-analysis reported 17 risk loci for IA (including both ruptured and unruptured cases) 7 . Another study further contextualized IA risk in relation to hypertension and smoking and refined the mapping of IA risk loci, underscoring shared cardiometabolic pathways and cell-type-specific programs 8 . However, the biological mechanisms linking these genetic variants to aneurysm formation and rupture remain poorly understood, and none of these discoveries have yet translated into preventative therapies. One reason may be the substantial genetic heterogeneity of IA, which complicates the identification of causal genes. Additionally, differences in IA prevalence across populations present challenges in replicating findings and generalizing risk prediction 9 , 10 . This gap in understanding, coupled with the absence of preventative pharmacological treatments, highlights the need for novel approaches to identify molecular biomarkers and therapeutic targets for IA and SAH. Proteins are central to numerous biological processes, and their dysregulation can lead to disease, making them key biomarkers for diagnosis, risk prediction, and understanding disease mechanisms 11 – 16 . Moreover, most therapeutic drugs target proteins 17 . Recent advances in large-scale proteomic profiling offer an unprecedented opportunity to assess the causal role of proteins in disease development, identify novel biomarkers, and uncover new therapeutic targets through integrative analyses with genomic data 18 – 21 . GWAS of plasma proteins have identified protein quantitative trait loci (pQTLs), enabling Mendelian randomization (MR) analyses to explore the causal impact of proteins on disease 22 – 27 . MR uses genetic variants as proxies for lifelong differences in exposure levels, in order to infer causal effects of the exposures on disease outcomes 28 , 29 . Complementary strategies, such as colocalization to reduce the risk of confounding, and triangulation with observational and rare variant data, can strengthen causal inference and target prioritization. Here, we conducted ancestry-specific GWAS meta-analyses of IA and SAH in European and East Asian populations, leveraging multiple large-scale resources, and then performed genetically stratified, proteome-wide two-sample MR using five proteomics cohorts to evaluate the effects of circulating proteins on aneurysm risk. Given that elevated blood pressure is a major causal risk factor for IA, we additionally applied GWAS-by-subtraction to partition systolic blood pressure (SBP)-related versus non-SBP genetic effects and then performed MR on the non-SBP-mediated IA component. We further prioritized signals using sensitivity analyses and colocalization, and triangulated evidence with observational associations in UK Biobank, rare variant gene burden tests in 426,295 exomes, and targeted analyses in a French-Canadian familial IA cohort. Collectively, these analyses prioritize a small set of circulating proteins as candidates for informing future mechanistic and therapeutic investigation. Methods Genome-wide association meta-analyses for IA and SAH We performed ancestry-specific meta-analyses of GWAS of IA and SAH across European and East Asian ancestries (see Data availability section). Within each ancestry (European or East Asian) and outcome (IA or SAH), we conducted fixed-effects inverse-variance weighted meta-analysis in METAL 30 . The individual summary statistics contributing to meta-analyses were derived from a combination of previously published GWAS datasets, as detailed in Supplementary Table 1 . Full details of sample sources and GWAS methodology for each contributing study are described in the original publications. In European ancestries, the IA meta-analysis ( n = 1,441,701; 16,594 cases; 1,425,107 controls) pooled Bakker 2020 Stage 1 7 , FinnGen non-ruptured cerebral aneurysm and SAH (release 12) 8 , and Million Veteran Program (MVP) cerebral aneurysm 28 . The SAH meta-analysis ( n = 982,264; 10,057 cases; 972,207 controls) combined Bakker 2020 SAH 7 , FinnGen SAH (release 12) 8 , and MVP SAH 31 . In East Asian ancestries, the IA meta-analysis ( n = 538,879; 6,418 cases; 532,461 controls) included Biobank Japan cerebral aneurysm 32 , Taiwan Precision Medicine Initiative (TPMI) intracranial hemorrhage 33 , 34 , China Kadoorie non-traumatic SAH 35 , and an Inuit cohort 36 . For analytical purposes, we grouped the Inuit cohort with the East Asian ancestry stratum because Inuit populations share substantial East Asian-related genetic ancestry owing to their demographic history 37 , 38 . The SAH meta-analysis combined Biobank Japan SAH 32 , TPMI intracranial hemorrhage 33 , 34 , and China Kadoorie nontraumatic SAH 35 ( n = 536,841; 4,447 cases; 532,394 controls). We performed post-analysis filtering following standard filtering criteria as previously detailed 31 , 39 . Briefly, we applied variant-level quality control to reduce heterogeneity and frequency instability. Single nucleotide polymorphisms (SNPs) showing evidence of between-study heterogeneity were excluded (Cochran’s Q-test P 80%). We then quantified minor allele frequency (MAF) variability per variant as MaxFreq - MinFreq and removed variants with MAF variability > 0.15 (> 15%) following standard filtering criteria 34 and that implemented previously https://github.com/huw-morris-lab/meta-analysis . The final meta-analyses thus retained SNPs with Q-test P ≥ 0.05, I 2 < 80%, and MAF variability ≤ 0.15. For each of the four meta-analyses, we identified linkage disequilibrium (LD)-independent association signals using LD clumping with a clumping window of 1 Mb, r 2 threshold of 0.001, and genome-wide significance level of 5 × 10 -8 . LD was computed against ancestry-matched reference panels. For European ancestry, we used a reference panel consisting of 50,000 randomly selected, unrelated individuals of European descent from the UK Biobank 40 (UKB 50k). For East Asian ancestry, the LD structure was derived from the East Asian subset of the 1000 Genomes Project 41 (1kGP EAS). We annotated the nearest gene for independent variants based on distance to the transcription start site. Proteomics study cohorts We leveraged four large-scale proteomics studies derived from individuals of European ancestry from ARIC 18 , deCODE 19 , Fenland 20 , and UKB-PPP 21 . Details of these studies can be found in their original publications and Supplementary Table 1 . Briefly, participants were profiled using the SomaScan v4 assay (SomaLogic) in the ARIC study comprising 4,657 circulating plasma proteins measured in up to 7,213 European American individuals, the deCODE study comprising 4,719 proteins measured in up to 35,559 Icelandic individuals, and the Fenland study comprising 4,775 proteins measured in up to 10,708 individuals. Participants in the UKB-PPP study were profiled on Olink Explore 3072 comprising 2,923 proteins measured in up to 34,557 individuals in the discovery cohort). In East Asian ancestries, we used GWAS from the East Asian subset of the UKB-PPP composed of GWAS of 2,923 proteins from 262 individuals and the Kyoto-Nagahama cohort encompassing 4,196 proteins (SomaScan v4) from 1,823 Japanese individuals. Genetic instrument selection We performed LD clumping using a clumping window of 1 Mb, r 2 threshold of 0.001, and genome-wide significance level of 5 × 10 -8 to identify independent variants from each of the four original studies above. Variants within 500 kb of the transcription start site of the protein-coding gene were considered cis -pQTLs. We used the same ancestry-matched reference panels described above (European: UKB 50k; East Asian: 1kGP EAS). Only variants with a MAF greater than 1% in these ancestry-matched panels were retained for further analysis. Two-sample Mendelian randomization We estimated the effect of circulating plasma protein levels on meta-analyzed IA outcomes using proteome-wide two-sample MR in a genetically stratified manner (TwoSampleMR v.0.5.7 42 ). MR relies on three instrumental-variable assumptions: (1) Relevance: the genetic instrument is associated with the exposure; (2) Independence: the instrument is independent of any confounders of the exposure-outcome relationship; and (3) Exclusion restriction: the instrument influences the outcome only through its effect on the exposure (i.e., no alternative causal pathways or horizontal pleiotropy). We excluded proteins within the major histocompatibility complex (MHC) due to the region’s complex linkage disequilibrium structure, which can confound genetic association signals and complicate interpretation of causal relationships 43 . For European analyses, we utilized proteomic GWAS summary data from ARIC 18 , deCODE 19 , Fenland 20 , and UKB-PPP 21 studies as exposures and assessed the effects of proteins in each cohort on European IA and SAH GWAS. For East Asian analyses, we used proteomic GWAS from the UKB-PPP 21 study as exposures and assessed the effect of proteins meta-analyzed East Asian IA and SAH outcomes. In this context, the term "protein" refers to the aptamer (SomaScan) or antibodies (Olink) targeting the respective protein. Summary statistics from protein GWAS (exposures) and IA outcome GWAS were harmonized using the harmonise_data() function. When a selected instrument was unavailable in the outcome dataset, we conducted a proxy SNP search using ancestry-matched LD reference panels from the instrument selection procedure. Proxy variants were identified using PLINK v.1.9 44 with parameters --ld-window=5000, --ld-window-kb=5000, --ld-window-r2=0.8. We retained proxies with MAFs ≤ 0.42. Harmonized results are shown in Supplementary Table 6-11 . Instrument strength was evaluated using F-statistics ( Supplementary Table 12 ), with values above 10 considered indicative of robust instruments and lower likelihood of weak instrument bias 45 , 46 . Causal estimates were derived using the mr() function and represent the odds ratio per 1 standard deviation increase in genetically predicted protein level. For proteins with one instrument, we applied the Wald ratio; for those with two or more instruments, we used an inverse-variance weighted random-effects model. We controlled for multiple testing by using a Benjamini-Hochberg false discovery rate (FDR) 47 threshold of 5%, consistent with prior studies 25 , 48 . However, to ensure stringency, we performed the correction within each proteomics cohort and at the single-trait level. MR analyses adhered to STROBE-MR reporting guidelines 28 , 29 ( Supplementary Note 1 ). GWAS-by-subtraction We conducted GWAS-by-subtraction 49 to identify genetic effects on intracranial aneurysm that were not fully mediated by SBP-related pathways. European-ancestry IA summary statistics ( n = 1,441,701; 16,594 cases; 1,425,107 controls) were obtained from the meta-analysis we conducted earlier. European-ancestry SBP GWAS summary statistics ( n = 757,601) were taken from Evangelou et al. 50 . Subtraction analyses were performed following the GWAS-by-subtraction framework in Genomic SEM 51 , using a European ancestry subset of the 1000 Genomes Project 41 as the LD reference panel. GWAS-by-subtraction partitions intracranial aneurysm into SBP-related effects (g SBP ) and non-SBP-mediated latent genetic effects (g nonSBP ) which represents IA genetic associations not fully mediated by SBP-related pathways ( see Data availability ). Genetic correlation analyses were performed using LDSC 52 . We performed MR analyses on the g nonSBP GWAS following the same procedure described previously for the original IA GWAS and used an FDR-adjusted P value threshold of 5% for significance. Sensitivity analyses with alternative MR methods To ensure the robustness of causal estimates, we applied multiple sensitivity checks, including heterogeneity testing, alternative MR methods (weighted median, weighted mode, MR-Egger), and Steiger directionality testing. For proteins with two or more instruments, we calculated a heterogeneity P value using Cochran’s Q (Q_pval) and I 2 . Associations with I 2 ≥ 0.5 and Q_pval < 0.05 were considered heterogeneous. For proteins with three or more instruments, we additionally applied weighted median, weighted mode, and MR-Egger methods. Consistency in effect direction across these methods was required. Directional pleiotropy was tested using mr_pleiotropy_test() (significant if P < 0.05). We also used Steiger filtering (directionality_test()) to exclude variants suggesting reverse causation. Colocalization We used the colocalization method, SharePro 53 , to test whether protein levels and GWAS outcomes share causal variants for associations passing sensitivity analyses above. Analyses were conducted within 1 Mb of the lead cis -pQTL using default priors. Colocalization was defined as a posterior probability (PP.H4) ≥ 0.8. Heterogeneity analyses To identify proteins contributing to distinct biological pathways for IA or SAH, we evaluated between-outcome heterogeneity between IA and SAH across four European cohorts: ARIC, deCODE, Fenland, and UKB-PPP. MR summary statistics were obtained per cohort, and proteins common to both outcomes were retained for comparison. For each shared protein, we conducted a fixed-effect meta-analysis using the rma() function in the metafor R package, combining MR effect estimates from IA SAH while weighting by the inverse of their variance. Heterogeneity was assessed using three statistics: (i) Cochran’s Q, which tests the null hypothesis that effect sizes are consistent across outcomes; (ii) the Q-test P value, where P < 0.05 indicates statistically significant heterogeneity; and (iii) I 2 , which quantifies the proportion of total variability attributable to heterogeneity rather than sampling error. Proteins showing significant heterogeneity ( P < 0.05) were interpreted as potential candidates for differential associations with IA and SAH. Observational association analyses for cerebrovascular diseases We used logistic regression to evaluate whether protein levels were associated with risk of cerebrovascular diseases in the UK Biobank 40 . Cerebrovascular disease was defined from linked hospital inpatient diagnoses (ICD-10 and ICD-9), death-registry codes, and self-reported data ( Supplementary Note 2 ). After excluding participants lacking complete covariate, proteomic, or diagnostic information, 442,896 individuals remained representing 30,315 cerebrovascular disease cases and 412,581 controls. Of these participants, 39,649 had overlapping Olink proteomics data (2,975 cerebrovascular disease cases and 36,674 controls) and were analyzed. Plasma protein abundances were measured with the Olink Explore 3072 panel. For each protein, NPX values were rank-based inverse-normal transformed. Logistic regression models adjusted for age at recruitment, sex, recruitment center, Olink measurement batch, Olink sample processing time, and genetic ancestry (using the first 10 genetic principal components). We assessed the nine unique proteins which achieved FDR-adjusted P < 0.05 from MR analyses. Of these proteins, six were present in the UK Biobank. Thus, statistical significance was assessed at P < 0.05 / 6 = 8.3×10 -3 to account for the number of unique proteins tested. Incident associations were derived from Deng et al.’s plasma proteome atlas in 53,026 UK Biobank participants, in whom 2,920 plasma proteins were measured at baseline and linked to 660 incident ICD-10 defined diseases over a median 14.8-year follow-up 54 . We queried the public Proteome-Phenome Atlas for hazard ratios and 95% confidence intervals per standard deviation higher protein level for pre-specified cerebrovascular, aneurysmal and vascular outcomes ( Supplementary Note 3 ). Rare variant gene-based association analysis in the UK Biobank We leveraged whole genome sequencing data from the UK Biobank, comprising of 426,295 participants. Cerebrovascular disease cases and controls were defined as in the observational analyses. To capture the potential effects of deleterious rare variants, for each MR-prioritized gene, we constructed six annotation variant masks and applied three allele-frequency bins (singletons; MAF ≤ 0.001; MAF ≤ 0.01). Masks were: M1 (loss of function, LoF only): high-confidence predicted loss-of-function. M2 (LoF + missense 5/5): LoF plus missense called damaging by all 5 in-silico predictors. M3 (LoF + any deleterious missense): LoF plus missense called damaging by ≥1/5 predictors (also includes 5/5). M4 (LoF + all missense): LoF plus all missense. M5 (all coding): LoF + all missense + synonymous. M6 (synonymous only): negative-control mask. Burden association tests were conducted using REGENIE 55 using a Firth logistic regression model, adjusting for age at recruitment, sex, the first ten genetic principal components, sequencing batch, and assessment center. Analyses were restricted to individuals of European ancestry to minimize bias due to population stratification. Rare exonic variant analysis in a French-Canadian cohort We analyzed exome sequencing data from a French-Canadian (FC) familial IA cohort ( n = 234; 32 cases from six families and 202 controls) using the recruitment, sequencing, alignment, variant-calling, and quality-control procedures described in our previous publication 9 . We focused on potential FC founder variants in nine MR-prioritized genes (MAF < 0.05 in any gnomAD v2.2.1 ancestry). Two-sided Fisher’s exact test was applied to test the variant carrier status in FC IA cases compared to FC controls, using the 2×2 table and reported the odds ratio with exact 95% confidence interval (CI). We used Bonferroni correction to account for multiple testing across all 13 variants identified across the nine genes ( P < 0.05 / 13 = 0.0038). Druggability assessment We evaluated the druggability of putatively causal proteins using the druggable genome by Finan et al. 56 , DrugBank (v5.1.12) 57 , and Open Targets (v24.03) 58 . Heatmaps were constructed using the pheatmap package. Data availability All contributing cohorts obtained ethical approval from their institutional ethics review boards, and all participants provided written informed consent. The contributing proteomics cohorts include the ARIC Study, deCODE study, Fenland study, UK Biobank, and Kyoto University Nagahama study. The UK Biobank has approval from the North West Multi-centre Research Ethics Committee as a Research Tissue Bank, and analyses were conducted under UK Biobank application 73958. Meta-analyzed IA GWAS (European and East Asian) and meta-analyzed SAH GWAS (European and East Asian) will be deposited on GWAS catalog upon publication of this study. The Inuit cohort included in the East Asian IA meta-analysis will be deposited on GWAS catalog upon publication of this study. The non-SBP-mediated IA GWAS from the GWAS-by-subtraction analysis will be deposited on GWAS catalog upon publication. Results Genome-wide association meta-analysis of intracranial aneurysm We performed ancestry-stratified genome-wide association meta-analysis leveraging data from seven large-scale biobank and case-control resources encompassing nearly 1.5 million individuals ( Figure 1 ). Analyses were conducted on IA and SAH in individuals of European and East Asian ancestry. Cohort details are provided in Supplementary Table 1 . In European ancestries, we identified 34 and 14 LD independent variants for IA and SAH, respectively ( Supplementary Table 2 and Supplementary Table 3 ). In East Asian ancestries, we found five and two LD independent variants for IA and SAH, respectively ( Supplementary Table 4 and Supplementary Table 5 ). These two loci ( AKAP8 and WIZ ) were novel associations for IA and SAH risk and are uniquely found in the East Asian populations. Download figure Open in new tab Figure 1. Genome-wide association results for intracranial aneurysm (IA) and subarachnoid hemorrhage (SAH) stratified by ancestry. Manhattan plots show association P values (-log 10 P ; y-axis) for variants across the autosomes (chromosomes 1-22; x-axis), with alternating colors denoting adjacent chromosomes. Meta-analyses were conducted separately in European (left) and East Asian (right) cohorts for IA (top row) and SAH (bottom row). Sample sizes are given in each panel: European IA ( n = 1,441,701); East Asian IA ( n = 538,879); European SAH ( n = 982,264); East Asian SAH ( n = 536,841). The horizontal dashed line marks the genome-wide significance threshold ( P = 5×10 -8 ). Each point represents a single variant, with taller peaks indicating stronger association signals. Lead variants in linkage disequilibrium-independent regions are labeled with the nearest transcription start site gene (clumping window 1 Mb, r 2 < 0.001). Proteome-wide Mendelian randomization (MR) Across both European ( Figure 2a ) and East Asian ancestries ( Figure 2b ), we tested 15,611 protein-disease associations in total (IA and SAH across ancestries) ( Table 1 ). We confirmed that the genetic instruments explained a sufficient proportion of variance in protein levels (F > 10) across cohorts in both ancestries ( Supplementary Table 12 ). To address multiple testing, we used an FDR P < 0.05 as the threshold for significance per cohort and identified a total of 12 FDR significant protein-disease associations (< 0.1% of the total tests). These associations were composed of 11 protein-disease pairs involving IA and a single association for SAH in European ancestries ( Figure 2c , Supplementary Figure 1, and Supplementary Table 13 ). The 12 FDR-adjusted significant protein-disease associations in European ancestry involved nine unique proteins, with eight associated with IA and one with SAH. For IA, risk-increasing associations were observed for SLMAP, AMBP, ENTPD6, and PLEKHA1, while protective associations were seen for SIRT2, JAG1, ADH4, and NAGLU. For SAH, only ADAM23 reached significance. The direction and magnitude of effects were generally consistent across cohorts. PLEKHA1 replicated in ARIC (OR = 1.29, 95% CI 1.16–1.44, P = 5.54×10 -6 ) and deCODE (OR = 1.22, 95% CI 1.12–1.33, P = 1.03×10 -5 ). ADH4 replicated in Fenland (OR = 0.56, 95% CI 0.44–0.71, P = 1.80×10 -6 ), UKB-PPP (OR = 0.62, 95% CI 0.51–0.76, P = 1.80×10 -6 ), and ARIC (OR = 0.86, 95% CI 0.81–0.91, P = 1.71×10 -7 ). In East Asian ancestries, no protein-disease associations were significant after FDR correction ( Supplementary Figure 1 and Supplementary Table 14 ). Download figure Open in new tab Figure 2. Identification of putatively causal proteins for intracranial aneurysms across ancestries. (A) European ancestry Mendelian randomization (MR) design. Two-sample MR analyses were performed using cis -pQTL instruments from four European proteomics cohorts: ARIC ( n = 7,213; 4,657 proteins), deCODE ( n = 35,559; 4,907 proteins), Fenland ( n = 10,708; 4,775 proteins), and UKB-PPP ( n = 34,557; 2,923 proteins) with outcomes of intracranial aneurysm (IA) and aneurysmal subarachnoid hemorrhage (SAH). (B) East Asian ancestry MR design. Conceptually identical MR analyses were carried out using instruments derived from two East Asian proteomics cohorts (UKB-PPP: n = 262; 2,923 proteins and Kyoto-Nagahama: n = 1,823; 4,196 proteins) for IA and SAH. (C) Significant European MR results (FDR-adjusted P < 0.05). Forest plots show odds ratios (ORs) per standard-deviation increase in genetically predicted circulating protein levels with 95% confidence intervals for IA (left) and SAH (right). Points represent cohort-specific MR estimates (colors denote proteomic cohorts: ARIC, deCODE, Fenland, UKB-PPP), and the vertical dashed line indicates the null (OR = 1). MR estimates and P values were obtained using the inverse-variance-weighted random-effects method when instruments included more than one variant, and the Wald ratio when the instrument comprised a single variant. Higher ORs indicate increased risk per higher protein level, whereas ORs < 1 indicate lower risk. Abbreviations: MR, Mendelian randomization; IA, intracranial aneurysm; SAH, subarachnoid hemorrhage; OR, odds ratio; FDR, false discovery rate. View this table: View inline View popup Download powerpoint Table 1. Summary of proteome-wide MR results across cohorts and ancestries. For each proteomics cohort and outcome the table reports: (i) the number of proteins with valid instruments tested (“MR number of tests”), (ii) the count reaching significance after multiple-testing correction (FDR-adjusted P < 0.05, controlled at the cohort-trait level), (iii) the count that also passed standard MR sensitivity checks (directionality/heterogeneity/pleiotropy; excluding colocalization), and (iv) the count additionally supported by colocalization (PP.H4 ≥ 0.8). European cohorts: ARIC, deCODE, Fenland, and UKB-PPP; East Asian cohorts: UKB-PPP and Kyoto-Nagahama. Abbreviations: MR, Mendelian randomization; IA, intracranial aneurysm; SAH, subarachnoid hemorrhage; FDR, false discovery rate; PP.H4, posterior probability of a shared causal variant. GWAS-by-subtraction Given prior evidence that elevated blood pressure, a hallmark of hypertension, causally increases the risk of IA 59 , 60 , we used a GWAS-by-subtraction approach 49 , 51 , 61 , 62 to disentangle the SBP-mediated effects of the MR-prioritized proteins in European ancestries. Upon GWAS-by-subtraction, genetic effects not mediated through SBP were less polygenic compared to genetic effects associated with SBP ( Figure 3a and Figure 3b ). We identified 1,640 genome-wide significant variants, 26 of which represented LD-independent loci ( Supplementary Table 15 ). Additionally, the genetic correlation between SBP and IA attenuated upon removal of SBP-associated genetic signals. Prior to removal, the original meta-analyzed European ancestry IA trait demonstrated a correlation of r g = 0.256 (standard error [SE] = 0.0326). Removing SBP-associated effects using GWAS-by-subtraction attenuated the genetic correlation estimates between SBP and non-SBP IA (g nonSBP ) to r g = 0.2132 (SE = 0.0324) while SBP showed high correlation with the SBP-related latent GWAS (g SBP ) with r g = 0.9995 (SE = 0.0408). Taken together, these findings suggest that non-SBP-mediated genetic signals were isolated effectively. Download figure Open in new tab Figure 3. Disentangling SBP-related and non-SBP-mediated genetic effects on intracranial aneurysm and follow-up proteome-wide MR. (A) Illustration of the GWAS-by-subtraction model for systolic blood pressure (SBP)-derived effects on intracranial aneurysm (IA) in European ancestry. 𝑔 SBP : SBP genetic latent factors 𝑔 nonSBP : non-SBP genetic latent factors : SNP effect on SBP genetic latent factor : SNP effect on non-SBP genetic latent factor : effect of SBP genetic latent factors on genetic components of SBP : effect of SBP genetic latent factors on genetic components of IA : effect of non-SBP genetic latent factors on genetic components of IA (B) Manhattan plots. Genome-wide association results for 𝑔 SBP (left, red) and 𝑔 nonSBP (right, blue). The y-axis shows -log 10 P and the x-axis shows autosomes 1-22 (alternating shades). The dashed line marks the genome-wide significance threshold ( P = 5×10 -8 ). (C) Significant proteins for non-SBP-mediated IA (FDR-adjusted P < 0.05). Forest plots display odds ratios (ORs) per s.d. increase in genetically predicted circulating protein levels with 95% confidence intervals (CIs) for IA, using cis -pQTL instruments from European proteomic cohorts (ARIC, Fenland, UKB-PPP). Proteins passing FDR correction are shown (PLEKHA1, ADH4, JAG1). Points represent cohort-specific estimates, and the vertical dashed line indicates the null (OR = 1). MR estimates and P values were computed using the inverse-variance-weighted random-effects method when instruments contained more than one variant and the Wald ratio for single-variant instruments. Abbreviations: IA, intracranial aneurysm; SBP, systolic blood pressure; MR, Mendelian randomization; OR, odds ratio; pQTL, protein quantitative trait locus; FDR, false discovery rate. Next, we applied proteome-wide MR to the non-SBP-mediated IA GWAS (g nonSBP ) in European anceestry and found that three of the eight MR-prioritized proteins for IA had consistent effects on IA that were not acting through SBP (FDR-adjusted P or q value < 0.05; Figure 3c and Supplementary Table 16 ). The most consistent signal was ADH4, which showed concordant protective effects across three cohorts: ARIC (OR = 0.92, 95% CI 0.90–0.95, q = 1.37×10 -3 ), Fenland (OR = 0.74, 0.65–0.84, q = 1.14×10 -2 ), and UKB-PPP (OR = 0.78, 95% CI 0.70–0.87, q = 1.27×10 -2 ). Similarly, higher JAG1 levels were protective (OR = 0.78, 95% CI 0.69–0.88, q = 1.89×10 -2 ), whereas higher circulating PLEKHA1 levels were associated with increased IA risk (OR = 1.14, 95% CI 1.07–1.21, q = 1.35×10 -2 ). Collectively, these results nominate ADH4 and JAG1 as putative non-SBP-mediated protective factors and PLEKHA1 as a risk-increasing factor for IA, warranting downstream validation. Sensitivity analyses with alternative MR methods The 12 FDR-adjusted significant protein-disease associations passed sensitivity analyses ( Supplementary Table 17 ). Sensitivity analyses with colocalization To further mitigate bias from LD, we performed colocalization analyses. Of the 12 FDR-significant protein-disease associations that passed all MR sensitivity checks, six (50%) were supported by colocalization evidence, mapping to three unique proteins ( Supplementary Table 17 ). For IA, ADH4 was supported by colocalization based on genetic associations in ARIC (PP.H4 = 1.00), Fenland (PP.H4 = 0.85), and UKB-PPP (PP.H4 = 0.97). PLEKHA1 was supported based on genetic associations in ARIC (PP.H4 = 0.92) and deCODE (PP.H4 = 0.93), and SLMAP was supported based on genetic associations in UKB-PPP (PP.H4 = 0.89), indicating that there was a shared causal variant underlying circulating protein levels and IA risk. Two additional targets showed moderate colocalization with the outcomes. SIRT2 with IA (PP.H4 = 0.78) and ADAM23 with SAH (PP.H4 = 0.68). Collectively, these findings strengthen ADH4 and PLEKHA1 as putative causal mediators of IA, while suggesting that the residual MR signals involving AMBP, ENTPD6, JAG1, and NAGLU may reflect LD with nearby but distinct variants or context-dependent regulation. Heterogeneity analyses between IA and SAH As an orthogonal analysis, we quantified between-outcome heterogeneity between IA and SAH MR estimates for shared proteins within each European ancestry cohort (ARIC, deCODE, Fenland, UKB-PPP). Across 6,680 IA-SAH comparisons, 58 (0.87%) showed nominal evidence of heterogeneity (Cochran’s Q P < 0.05), with I 2 = 74.1–92.8% ( Supplementary Table 18 ). The largest discordance was observed for SLMAP in UKB-PPP ( I 2 = 92.8%; Q = 13.93; P = 1.90×10 -4 ), indicating markedly different IA versus SAH effects within that cohort. In contrast, the median I 2 across all comparisons was 0%, consistent with low between-outcome discordance for most proteins. Together, these findings nominate a small subset of proteins with divergent associations across IA formation and rupture, suggesting partially distinct pathophysiologic pathways. Observational association analyses with cerebrovascular diseases To further examine whether the MR-identified proteins contribute to organ-specific disease mechanisms, we conducted observational association analyses for cerebrovascular disease in the UK Biobank (30,315 cases and 412,581 controls) using logistic regression for the nine proteins which achieved FDR-adjusted P < 0.05 in the MR analyses. Six of these proteins were present in the UK Biobank and could be assessed, of which four showed significant associations in the observational analyses. Notably, results were directionally concordant between MR and observational estimates for three proteins: SLMAP (logistic regression OR = 1.09, P = 3.04×10 -5 ), AMBP (OR = 1.26, P = 1.61×10 - 28 ), and ENTPD6 (OR = 1.07, P = 9.50×10 -4 ), whereas ADH4 showed a discordant association (OR = 1.12, P = 1.27×10 -7 ) ( Supplementary Table 19 ). Using Cox proportional hazards model analyses in UK Biobank from Deng et al. 54 , we next evaluated whether the six IA-prioritized proteins were associated with risk of incident cerebrovascular, aneurysmal, and vascular diseases ( Supplementary Table 20 ). AMBP showed the largest and most pervasive risk elevations, particularly for coronary, peripheral arterial, hypertensive renal and aneurysmal outcomes. SLMAP and ENTPD6 displayed broadly consistent positive associations, in line with MR and logistic regression findings. ADH4, SIRT2 and ADAM23 showed more modest or selective patterns across coronary, cerebrovascular and hypertensive renal diseases. Taken together, these prospective analyses indicate that several IA-prioritized proteins—particularly AMBP, SLMAP and ENTPD6—are associated with heightened risk of a broad range of incident cerebrovascular, aneurysmal and systemic vascular diseases in UK Biobank, supporting a shared vascular aetiological component linking these proteins to intracranial aneurysm. While JAG1, NAGLU, and PLEKHA1 were unable to be assessed, we undertook a pathway-proxy analysis in the UK Biobank by testing soluble interactors of JAG1 that were quantified in plasma as an example. Since JAG1 is a membrane-tethered ligand that is not reliably detected as a soluble analyte in the UK Biobank proteomic panels, we leveraged circulating, soluble interactors of JAG1 as pathway proxies. Among the five soluble interactors (VEGFA, DLL1, DLL4, ADAM17, and NOTCH1), only ADAM17 was not available in the UK Biobank. Logistic regression analyses suggested that higher DLL1 and VEGFA concentrations were associated with increased odds of cerebrovascular disease DLL1: OR = 1.24, P = 2.96×10 - 25 ; VEGFA: OR = 1.20, P = 1.67×10 - 19 ), both surpassing the Bonferroni threshold for four tests (α = 0.0125). By contrast, DLL4 and soluble NOTCH1 showed no evidence of association (DLL4: OR = 1.02, P = 0.24; NOTCH1: OR = 0.98, P = 0.38). These findings support a role for soluble components of the JAG1-NOTCH/angiogenic pathway in cerebrovascular disease risk and suggest complex, context-dependent regulation. Rare variant gene burden testing in 426,295 UK Biobank individuals for cerebrovascular disease We analyzed whole genome sequencing data from 426,295 UK Biobank participants to perform rare variant burden tests for cerebrovascular disease in the nine MR-prioritized genes. JAG1 showed the strongest nominal signal under the M2 mask (LoF and 5/5 deleterious missense) at MAF ≤ 0.01 and ≤ 0.001 (OR = 1.28, P = 0.021 for both), suggesting that carriers of rare, predicted-damaging JAG1 variants are at increased risk of cerebrovascular disease. This is directionally concordant with MR estimates, which implied that greater JAG1 activity is protective ( Supplementary Table 21 ). Additional nominal associations included PLEKHA1 (M6 singleton, OR = 4.27, P = 0.041), ADH4 (M5 MAF ≤ 0.01 and ≤ 0.001, OR = 0.86, P = 0.052 for both associations), and ENTPD6 (M5 MAF ≤ 0.01, OR = 1.07, P = 0.053). As expected for a negative control, M6 (synonymous-only) masks were broadly null, suggesting that the isolated nominal PLEKHA1 singleton finding likely reflects noise. Evidence of prioritized IA genes in a French-Canadian cohort We further assessed the contribution of potential French-Canadian (FC) founder variants in the nine MR-prioritized protein-coding genes using an FC familial IA cohort ( n = 234; 32 cases and 202 controls). After excluding variants with MAF ≥ 0.05 in gnomAD v2.1.1, we identified 13 variants. We performed two-sided Fisher’s exact tests to test the carrier state for each variant. Among the variants evaluated, ENTPD6 c.G67A:p.G23S (chr20:252070888, GRCh38) showed a significant case-control difference ( P = 0.0021) ( Supplementary Table 22 ). The variant was present in 8 of 32 FC IA patients (25.0%) versus 5 of 106 genotyped FC controls (4.7%). Interpreted as an enrichment in affected families, this corresponds to an odds ratio of 6.61 (95% CI 1.73–28.09), indicating substantially higher odds of carriage among affected familial IA patients. All other tested variants showed no statistically significant differences between affected and control families (two-sided P ≥ 0.19). Druggability assessment We next assessed the druggability of the nine MR-prioritized proteins by triangulating data from the druggable genome from Finan et al. 56 , DrugBank 57 , and Open Targets 58 ( Supplementary Table 23 and Figure 4 ). The druggable genome tiering stratified the set into one Tier 1 enzyme (ADH4), one Tier 2 enzyme (SIRT2), five Tier 3 membrane/secreted proteins (JAG1, AMBP, ADAM23, ENTPD6, NAGLU), and two unclassified targets (PLEKHA1, SLMAP). This suggests likely modality with small-molecule tractability for ADH4 and SIRT2. Concordantly, DrugBank contained bio-entity information for four of nine proteins, namely ADH4, SIRT2, NAGLU, and PLEKHA1, signaling the presence of sequence or cross-reference and, occasionally, ligand or structure information. In Open Targets, there exists no approved or clinically advanced therapeutic under development for these proteins. Download figure Open in new tab Figure 4. Druggability assessment of MR-prioritized proteins. Heat map summarizes proteins showing putative causal effects on intracranial aneurysm (IA) or aneurysmal subarachnoid hemorrhage (SAH) by two-sample MR (FDR-adjusted P < 0.05). Rows list proteins (ADH4, SIRT2, NAGLU, JAG1, AMBP, ADAM23, ENTPD6, PLEKHA1, SLMAP). Columns to the left provide target-annotation features: DrugBank (Yes/No), Open Targets (Yes/No), and druggability tier (Tier 1-3 or Unclassified). IA and SAH columns on the right display MR effects, with color indicating direction (red, risk-increasing; blue, risk-decreasing) and shade reflecting magnitude of the z score (MR estimate / standard error). MR estimates were derived using inverse-variance-weighted random-effects models for instruments with more than one variant and the Wald ratio for single-variant instruments. When a protein-disease had multiple associations, the effect estimates were averaged. Effect estimates are capped between -10 < z < 10 for visualization. Abbreviations: IA, intracranial aneurysm; SAH, subarachnoid hemorrhage; MR, Mendelian randomization; FDR, false discovery rate. Discussion In this study, we conducted meta-analyses of IA and SAH across European and East Asian ancestries. By integrating large-scale genetic and proteomic data from five independent cohorts, we identified eight circulating proteins linked to IA risk, and one linked to SAH using MR. We further strengthened these findings through comprehensive sensitivity and colocalization analyses, investigated a GWAS-by-subtraction framework that removed genetic effects acting through blood pressure, and assessed heterogeneity between IA and SAH. We then triangulated evidence across complementary analyses and datasets, including UK Biobank observational data, rare variant burden testing, and results from a FC familial cohort. A summary of the evidence is provided in Table 2 . These findings prioritise putatively causal proteins and pathways for mechanistic follow-up and therapeutic development. View this table: View inline View popup Download powerpoint Table 2. Triangulated evidence for circulating proteins implicated in intracranial aneurysm (IA) and subarachnoid hemorrhage (SAH) in European ancestries. For each protein, the table reports: (i) the number of European proteomics cohorts (of four) with a Mendelian randomization (MR) association at FDR-adjusted P < 0.05 and the effect direction (increasing risk vs. protective); (ii) the number of cohorts in which the MR association remained after removing genetic effects acting through systolic blood pressure (non-SBP-mediated IA from GWAS-by-subtraction); (iii) whether the signal passed MR sensitivity analyses and colocalization (PP.H4 ≥ 0.8) consistent with a shared causal variant; (iv) concordant or discordant observational associations with prevalent cerebrovascular disease in UK Biobank (logistic regression); (v) prospective associations with incident cerebrovascular, aneurysmal and other vascular diseases in UK Biobank (Cox proportional hazards models); (vi) gene based rare variant burden tests in UK Biobank exomes; and (vii) rare variant evidence in a French-Canadian familial IA cohort. Green entries indicate supportive/concordant evidence; red entries indicate discordant or negative evidence. “Nominal significance” denotes P < 0.05 without multiple-testing correction. Abbreviations: IA, intracranial aneurysm; SAH, subarachnoid hemorrhage; MR, Mendelian randomization; FDR, false discovery rate; SBP, systolic blood pressure; PP.H4, posterior probability that the pQTL and outcome GWAS signals colocalize (indicating a shared causal variant); LoF, loss-of-function; NA, not available since protein not measured; NS, not significant; M2 mask: LoF + missense predicted damaging by all five in-silico tools. *Results derived from Deng et al. 54 . As a Tier 1 druggable enzyme, ADH4 emerged as the most robust and translationally relevant signal in our analyses. Across ARIC, Fenland, and UKB-PPP, higher genetically predicted ADH4 levels were associated with lower IA risk. These effects persisted after removing SBP-mediated components, and the pQTL signal colocalized with the IA locus in all three cohorts (PP.H4 = 1.00 for all). Interestingly, the UK Biobank observational association was directionally discordant for cerebrovascular disease, which may reflect residual confounding or differences in outcome definition. Because ADH4 is an alcohol dehydrogenase, we considered whether its association might operate via ethanol metabolism. However, class I alcohol dehydrogenases (ADH1A, ADH1B, ADH1C) are more central to hepatic ethanol oxidation at physiological concentrations, and in our MR analyses all three—ADH1A, ADH1B, and ADH1C—showed null effects on IA risk. Mechanistically, ADH4 catalyzes retinol oxidation, contributing to retinoic-acid synthesis—a pathway with vascular anti-inflammatory and matrix-stabilizing properties consistent with a protective IA effect 63 , 64 . In vivo, retinoic acid attenuates aneurysm progression by slowing angiotensin II-driven abdominal aortic aneurysm growth 65 , reducing vascular inflammation and matrix-metalloproteinase (MMP) activity 66 , and limiting stenosis by suppressing vascular smooth-muscle-cell migration and MMP-9 67 . Taken together, these observations suggest that the ADH4 signal is more consistent with a retinoid-centric mechanism than with ethanol clearance per se; accordingly, while reduced alcohol consumption remains a relevant public-health consideration, our genetic results point to higher ADH4, and thus greater retinoic-acid flux, as a plausible causal pathway that may stabilize the arterial wall. PLEKHA1 showed a consistent risk-increasing association with IA, supported by significant MR estimates across two cohorts that persisted after removing SBP-related effects and demonstrated consistent colocalization. Biologically, PLEKHA1 encodes a pleckstrin-homology-domain adaptor that binds PI(3,4)P 2 at the plasma membrane and scaffolds phosphoinositide/PI3K signaling 68 , 69 . This axis regulates endothelial survival, permeability, and angiogenesis 70 which are processes central to vessel-wall remodeling in aneurysm pathophysiology 71 . Although PLEKHA1 is best known from the age-related macular degeneration locus 72 , its role within VEGF-PI3K-AKT pathways 73 provides a coherent mechanism by which higher circulating PLEKHA1 could promote pro-angiogenic and inflammatory changes in the cerebral arterial wall. JAG1 showed protective effects on IA risk and may act through non-SBP-mediated genetic effects. However, it colocalization evidence was equivocal. Our gene burden analysis showed nominal significance of damaging JAG1 variants, and pathway-proxy analysis showed that soluble JAG1 interactors, including DLL1 and VEGFA, were significantly associated with cerebrovascular disease in the UK Biobank, consistent with Notch/angiogenic signaling in the vessel wall. These findings align with previous studies establishing Jagged1–Notch as essential for vascular smooth-muscle recruitment and arterial integrity 74 , 75 . By contrast, SLMAP showed both colocalization based on genetic association and concordant observational association in the UK Biobank. SLMAP has been linked to endothelial dysfunction in animal resistance vessels 76 and to human microvascular disease in association studies 77 , but not specifically to human resistance-vessel dysfunction or to increased IA risk. We also identified extracellular matrix (ECM)-linked proteins such as AMBP, NAGLU, and ADAM23 that were prioritized through MR but did not show evidence of colocalization. Given the importance of ECM disruption in IA pathophysiology, these proteins could plausibly link circulating protein levels to arterial wall integrity. Nevertheless, further investigation is warranted to clarify the mechanisms underlying these associations. This study has several strengths. First, we included large, ancestry-stratified GWAS meta-analyses and integrated them with large-scale proteomics datasets, enabling replication across independent cohorts, two measurement platforms (SomaScan and Olink), and two ancestral groups. Second, we conducted MR with stringent instrument selection, extensive sensitivity analyses, and colocalization to ensure causal validity. Third, we triangulated genetic with observational data and used GWAS-by-subtraction to derive IA components whose effects were not fully mediated by blood pressure alongside other orthogonal analyses to refine mechanisms and highlight biologically plausible targets. Our study also has limitations. First, the signal specific to aneurysmal SAH was limited, despite efforts to increase power through GWAS meta-analyses. In European ancestry analyses, only one protein reached significance from MR for SAH, while no significant association was observed in East Asian ancestry analyses. Thus, conclusions regarding rupture phenotypes remain tentative and require further validation in larger GWAS. Second, the proteomic coverage, although broad, remains incomplete, and relevant proteins may not have been assayed. Third, even among the measured proteins, some lacked sufficiently strong or available genetic instruments, reducing power and precluding evaluation. Last, MR reflects lifelong, genetically proxied differences in protein levels and may not capture acute changes surrounding IA formation or rupture. In conclusion, integrating the largest GWAS of IA across two ancestries with high-resolution population-scale plasma proteomics, we prioritized a set of proteins that appear to causally influence IA formation and rupture. These proteins provide mechanistic biomarkers that inform insights into IA pathophysiology and represent actionable entry points for future therapeutic development. Code availability We used R v4.1.2 ( https://www.r-project.org/ ), TwoSampleMR v.0.5.6 ( https://mrcieu.github.io/TwoSampleMR/ ), SharePro v5.0 ( https://github.com/zhwm/SharePro_coloc ) PLINK v1.9 ( http://pngu.mgh.harvard.edu/purcell/plink/ ), GenomicSEM ( https://github.com/GenomicSEM/GenomicSEM ) Data Availability All contributing cohorts obtained ethical approval from their institutional ethics review boards, and all participants provided written informed consent. - The contributing proteomics cohorts include the ARIC Study, deCODE study, Fenland study, UK Biobank, and Kyoto University Nagahama study. - The UK Biobank has approval from the North West Multi-centre Research Ethics Committee as a Research Tissue Bank, and analyses were conducted under UK Biobank application 73958. - Meta-analyzed IA GWAS (European and East Asian) and meta-analyzed SAH GWAS (European and East Asian) will be deposited on GWAS catalog upon publication of this study. - The Inuit cohort included in the East Asian IA meta-analysis will be deposited on GWAS catalog upon publication of this study. - The non-SBP-mediated IA GWAS from the GWAS-by-subtraction analysis will be deposited on GWAS catalog upon publication. Author contributions Conception and design: C.-Y.S., S.Z. Methodology: C.-Y.S., S.Z. Data curation: C.-Y.S., J.M., C.L., S.Z. Data Analysis: C.-Y.S. Visualization: C.-Y.S. Writing—Original Draft: C.-Y.S. Writing—Review and Editing: All authors Supervision: S.Y., S.Z. Project administration: S.Z. Funding acquisition: S.Z. Competing Interests The authors declare no competing interests. Download figure Open in new tab Supplementary Figure 1. Cohort-specific volcano plots for proteome-wide MR across ancestries. (A) European cohorts (ARIC, deCODE, Fenland, UKB-PPP) and (B) East Asian cohorts (UKB-PPP, Kyoto-Nagahama). Each panel shows the MR odds ratio (OR) per s.d. increase in genetically predicted circulating protein level (x-axis) versus -log 10 P (y-axis) for intracranial aneurysm (IA) and aneurysmal subarachnoid hemorrhage (SAH). Points represent individual proteins; red points (“Pass”) meet the multiple-testing threshold (FDR-adjusted P 1 indicates higher risk per higher protein level; OR < 1 indicates lower risk. MR estimates and P values were derived using inverse-variance-weighted random-effects models for instruments with more than one variant and the Wald ratio for single-variant instruments. Abbreviations: MR, Mendelian randomization; IA, intracranial aneurysm; SAH, subarachnoid hemorrhage; OR, odds ratio; FDR, false discovery rate. View this table: View inline View popup Supplementary Note 1: STROBE-MR checklist of recommended items to address in reports of Mendelian randomization studies 1 2 Note: Page number will be added at the proof-reading stage. View this table: View inline View popup Download powerpoint Supplementary Note 2. Cerebrovascular disease case definition codes in the UK Biobank. Field IDs and coding used for case/control definition Supplementary Note 3. Incident vascular disease phenotypes considered in UK Biobank proteome-phenome analyses For the Cox proportional hazards analyses, we queried Deng et al.’s Proteome-Phenome Atlas for all incident endpoints available for the six IA-prioritised proteins. Among the 660 incident ICD-10-defined diseases in the atlas, 61 unique vascular and cerebrovascular phenotypes were returned at least once. We then pre-classified these into two tiers based on their relevance to intracranial aneurysm and cerebrovascular pathology. Tier 1 (core cerebrovascular/arterial phenotypes) Direct cerebrovascular or arterial disease outcomes and aneurysm phenotypes: Stroke, including SAH; Stroke, excluding SAH; Ischaemic stroke, excluding all haemorrhages; Transient ischaemic attack; Sequelae of cerebrovascular disease; Other specified cerebrovascular diseases, other cerebrovascular disorders in diseases classified elsewhere; Cerebral atherosclerosis; Aortic aneurysm; Abdominal aortic aneurysm (AAA); Other aneurysm; Peripheral artery disease; Other peripheral vascular diseases; Other diseases of arteries and capillaries . Tier 2 (upstream drivers / systemic atherosclerotic phenotypes) Vascular risk factor and systemic atherosclerotic outcomes that are biologically adjacent to IA and cerebrovascular disease and were used to provide broader vascular context: Hypertension; Hypertension, essential; Hypertensive renal disease; Atherosclerosis, excluding cerebral and coronary sclerosis; Ischaemic heart disease, wide definition; Coronary atherosclerosis; Angina pectoris; Unstable angina pectoris; Major coronary heart disease event; Myocardial infarction, strict; Myocardial infarction, with ST-elevation; Myocardial infarction, without ST-elevation; Arterial embolism and thrombosis of lower extremity artery; Other arterial embolism and thrombosis . Incident hazard ratios and 95% confidence intervals for these Tier 1-2 outcomes were extracted for each of the six IA-prioritised proteins and summarised in Table 2 and Supplementary Table 20 . Acknowledgments C.-Y.S. is supported by a CIHR Canada Graduate Scholarship Doctoral Award (Funding Reference Number: 187673), an FRQS doctoral training scholarship, and a Lady Davis Institute/TD-Bank Scholarship. M.H. is supported by the Japan Student Services Organization (Graduate Scholarship for Degree-Seeking Study Abroad) and by the Watanabe Foundation (6th Toshizo Watanabe International Scholarship). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. References 1. ↵ Intracranial Aneurysms | New England Journal of Medicine . https://www.nejm.org/doi/full/10.1056/NEJM199701023360106#sec-1 . 2. ↵ Vlak , M. H. , Algra , A. , Brandenburg , R. & Rinkel , G. J . Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: a systematic review and meta-analysis . The Lancet Neurology 10 , 626 – 636 ( 2011 ). OpenUrl PubMed 3. ↵ GBD 2021 Global Subarachnoid Hemorrhage Risk Factors Collaborators. Global, Regional, and National Burden of Nontraumatic Subarachnoid Hemorrhage: The Global Burden of Disease Study 2021 . JAMA Neurol 82 , 765 – 787 ( 2025 ). OpenUrl PubMed 4. ↵ Rooij , N. K. de , Linn , F. H. H. , Plas , J. A. van der , Algra , A. & Rinkel , G. J. E. Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends . Journal of Neurology, Neurosurgery & Psychiatry 78 , 1365 – 1372 ( 2007 ). OpenUrl Abstract / FREE Full Text 5. ↵ Ruigrok , Y. M. , Veldink , J. H. & Bakker , M. K . Drug classes affecting intracranial aneurysm risk: Genetic correlation and Mendelian randomization . European Stroke Journal 9 , 687 – 695 ( 2024 ). OpenUrl PubMed 6. ↵ Chalouhi , N. , Hoh , B. L. & Hasan , D . Review of Cerebral Aneurysm Formation, Growth, and Rupture . Stroke 44 , 3613 – 3622 ( 2013 ). OpenUrl FREE Full Text 7. ↵ Bakker , M. K. et al. Genome-wide association study of intracranial aneurysms identifies 17 risk loci and genetic overlap with clinical risk factors . Nat Genet 52 , 1303 – 1313 ( 2020 ). OpenUrl CrossRef PubMed 8. ↵ Adkar , S. S. et al. Dissecting the Genetic Architecture of Intracranial Aneurysms . Circulation: Genomic and Precision Medicine 18 , e004626 ( 2025 ). OpenUrl 9. ↵ Zhou , S. , et al. RNF213 Is Associated with Intracranial Aneurysms in the French-Canadian Population . The American Journal of Human Genetics 99 , 1072 – 1085 ( 2016 ). OpenUrl CrossRef PubMed 10. ↵ Zhou , S. et al. Genome-wide association analysis identifies new candidate risk loci for familial intracranial aneurysm in the French-Canadian population . Sci Rep 8 , 4356 ( 2018 ). OpenUrl PubMed 11. ↵ Butler-Laporte , G. et al. The dynamic changes and sex differences of 147 immune-related proteins during acute COVID-19 in 580 individuals . Clinical Proteomics 19 , 34 ( 2022 ). 12. Paranjpe , I. et al. Proteomic characterization of acute kidney injury in patients hospitalized with SARS-CoV2 infection . Commun Med 3 , 1 – 10 ( 2023 ). OpenUrl CrossRef PubMed 13. Arrell , D. K. , Neverova , I. & Van Eyk , J. E. Cardiovascular Proteomics . Circulation Research 88 , 763 – 773 ( 2001 ). OpenUrl Abstract / FREE Full Text 14. Anderson , N. L. & Anderson , N. G . The Human Plasma Proteome: History, Character, and Diagnostic Prospects* . Molecular & Cellular Proteomics 1 , 845 – 867 ( 2002 ). OpenUrl PubMed 15. Su , C.-Y. et al. Circulating proteins to predict COVID-19 severity . Sci Rep 13 , 6236 ( 2023 ). OpenUrl PubMed 16. ↵ Carrasco-Zanini , J. et al. Proteomic signatures improve risk prediction for common and rare diseases . Nat Med 30 , 2489 – 2498 ( 2024 ). OpenUrl CrossRef PubMed 17. ↵ Santos , R. et al. A comprehensive map of molecular drug targets . Nat Rev Drug Discov 16 , 19 – 34 ( 2017 ). OpenUrl CrossRef PubMed 18. ↵ Zhang , J. et al. Plasma proteome analyses in individuals of European and African ancestry identify cis-pQTLs and models for proteome-wide association studies . Nat Genet 54 , 593 – 602 ( 2022 ). OpenUrl CrossRef PubMed 19. ↵ Ferkingstad , E. et al. Large-scale integration of the plasma proteome with genetics and disease . Nat Genet 53 , 1712 – 1721 ( 2021 ). OpenUrl CrossRef PubMed 20. ↵ Pietzner , M. et al. Mapping the proteo-genomic convergence of human diseases . Science 374 , eabj1541 ( 2021 ). 21. ↵ Sun , B. B. et al. Plasma proteomic associations with genetics and health in the UK Biobank . Nature 622 , 329 – 338 ( 2023 ). OpenUrl CrossRef PubMed 22. ↵ Selber-Hnatiw , S. et al. Phenome-wide Mendelian randomization identifying circulating proteins for cardiovascular traits in populations of African ancestry . 2025.06.10.25329388 Preprint at doi: 10.1101/2025.06.10.25329388 ( 2025 ). OpenUrl Abstract / FREE Full Text 23. Zheng , J. et al. Phenome-wide Mendelian randomization mapping the influence of the plasma proteome on complex diseases . Nat Genet 52 , 1122 – 1131 ( 2020 ). OpenUrl CrossRef PubMed 24. Yoshiji , S. et al. Integrative proteogenomic analysis identifies COL6A3-derived endotrophin as a mediator of the effect of obesity on coronary artery disease . Nat Genet 1 – 13 ( 2025 ) doi: 10.1038/s41588-024-02052-7 . OpenUrl CrossRef 25. ↵ Su , C.-Y. et al. Multi-ancestry proteome-phenome-wide Mendelian randomization offers a comprehensive protein-disease atlas and potential therapeutic targets . 2024.10.17.24315553 Preprint at doi: 10.1101/2024.10.17.24315553 ( 2024 ). OpenUrl Abstract / FREE Full Text 26. Yoshiji , S. et al. Proteome-wide Mendelian randomization implicates nephronectin as an actionable mediator of the effect of obesity on COVID-19 severity . Nat Metab 5 , 248 – 264 ( 2023 ). OpenUrl PubMed 27. ↵ Zhou , S. et al. A Neanderthal OAS1 isoform protects individuals of European ancestry against COVID-19 susceptibility and severity . Nat Med 27 , 659 – 667 ( 2021 ). OpenUrl PubMed 28. ↵ Skrivankova , V. W. et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration . BMJ 375 , n2233 ( 2021 ). 29. ↵ Skrivankova , V. W. et al. Strengthening the Reporting of Observational Studies in Epidemiology Using Mendelian Randomization: The STROBE-MR Statement . JAMA 326 , 1614 – 1621 ( 2021 ). OpenUrl CrossRef PubMed 30. ↵ Willer , C. J. , Li , Y. & Abecasis , G. R . METAL: fast and efficient meta-analysis of genomewide association scans . Bioinformatics 26 , 2190 – 2191 ( 2010 ). OpenUrl CrossRef PubMed Web of Science 31. ↵ Verma , A. et al. Diversity and scale: Genetic architecture of 2068 traits in the VA Million Veteran Program . Science 385 , eadj1182 ( 2024 ). 32. ↵ Sakaue , S. et al. A cross-population atlas of genetic associations for 220 human phenotypes . Nat Genet 53 , 1415 – 1424 ( 2021 ). OpenUrl CrossRef PubMed 33. ↵ Yang , H.-C. et al. The Taiwan Precision Medicine Initiative provides a cohort for large-scale studies . Nature 1 – 34 ( 2025 ) doi: 10.1038/s41586-025-09680-x . OpenUrl CrossRef 34. ↵ Chen , H.-H. et al. Population-specific polygenic risk scores for people of Han Chinese ancestry . Nature 1 – 10 ( 2025 ) doi: 10.1038/s41586-025-09350-y . OpenUrl CrossRef 35. ↵ Walters , R. G. et al. Genotyping and population characteristics of the China Kadoorie Biobank . Cell Genomics 3 , ( 2023 ). 36. ↵ Zhou , S. et al. Genetic architecture and adaptations of Nunavik Inuit . Proceedings of the National Academy of Sciences 116 , 16012 – 16017 ( 2019 ). OpenUrl Abstract / FREE Full Text 37. ↵ Flegontov , P. et al. Palaeo-Eskimo genetic ancestry and the peopling of Chukotka and North America . Nature 570 , 236 – 240 ( 2019 ). OpenUrl CrossRef PubMed 38. ↵ Raghavan , M. et al. The genetic prehistory of the New World Arctic . Science 345 , 1255832 ( 2014 ). 39. ↵ Iwaki , H. et al. Genomewide association study of Parkinson’s disease clinical biomarkers in 12 longitudinal patients’ cohorts . Movement Disorders 34 , 1839 – 1850 ( 2019 ). OpenUrl CrossRef PubMed 40. ↵ Bycroft , C. et al. The UK Biobank resource with deep phenotyping and genomic data . Nature 562 , 203 – 209 ( 2018 ). OpenUrl CrossRef PubMed 41. ↵ Auton , A. et al. A global reference for human genetic variation . Nature 526 , 68 – 74 ( 2015 ). OpenUrl CrossRef PubMed 42. ↵ Hemani , G. et al. The MR-Base platform supports systematic causal inference across the human phenome . eLife 7 , e34408 ( 2018 ). OpenUrl CrossRef PubMed 43. ↵ Butler-Laporte , G. et al. HLA allele-calling using multi-ancestry whole-exome sequencing from the UK Biobank identifies 129 novel associations in 11 autoimmune diseases . Commun Biol 6 , 1 – 17 ( 2023 ). OpenUrl PubMed 44. ↵ Purcell , S. et al. PLINK: A Tool Set for Whole-Genome Association and Population-Based Linkage Analyses . The American Journal of Human Genetics 81 , 559 – 575 ( 2007 ). OpenUrl CrossRef PubMed 45. ↵ Pierce , B. L. , Ahsan , H. & VanderWeele , T. J . Power and instrument strength requirements for Mendelian randomization studies using multiple genetic variants . International Journal of Epidemiology 40 , 740 – 752 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 46. ↵ Lawlor , D. A. , Harbord , R. M. , Sterne , J. A. C. , Timpson , N. & Davey Smith , G . Mendelian randomization: Using genes as instruments for making causal inferences in epidemiology . Statistics in Medicine 27 , 1133 – 1163 ( 2008 ). OpenUrl CrossRef PubMed 47. ↵ Benjamini , Y. & Hochberg , Y . Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing . Journal of the Royal Statistical Society. Series B (Methodological ) 57 , 289 – 300 ( 1995 ). OpenUrl CrossRef PubMed Web of Science 48. ↵ Zhao , H. et al. Proteome-wide Mendelian randomization in global biobank meta-analysis reveals multi-ancestry drug targets for common diseases . Cell Genomics 2 , 100195 ( 2022 ). 49. ↵ Demange , P. A. et al. Investigating the genetic architecture of noncognitive skills using GWAS-by-subtraction . Nat Genet 53 , 35 – 44 ( 2021 ). OpenUrl CrossRef PubMed 50. ↵ Evangelou , E. et al. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits . Nat Genet 50 , 1412 – 1425 ( 2018 ). OpenUrl CrossRef PubMed 51. ↵ Grotzinger , A. D. et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits . Nat Hum Behav 3 , 513 – 525 ( 2019 ). OpenUrl PubMed 52. ↵ Bulik-Sullivan , B. K. et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies . Nat Genet 47 , 291 – 295 ( 2015 ). OpenUrl CrossRef PubMed 53. ↵ Zhang , W. et al. SharePro: an accurate and efficient genetic colocalization method accounting for multiple causal signals . 2023.07.24.550431 Preprint at doi: 10.1101/2023.07.24.550431 ( 2023 ). OpenUrl Abstract / FREE Full Text 54. ↵ Deng , Y.-T. et al. Atlas of the plasma proteome in health and disease in 53,026 adults . Cell 188 , 253 – 271 .e7 ( 2025 ). OpenUrl CrossRef PubMed 55. ↵ Mbatchou , J. et al. Computationally efficient whole-genome regression for quantitative and binary traits . Nat Genet 53 , 1097 – 1103 ( 2021 ). OpenUrl CrossRef PubMed 56. ↵ Finan , C. et al. The druggable genome and support for target identification and validation in drug development . Science Translational Medicine 9 , eaag1166 ( 2017 ). 57. ↵ Wishart , D. S. et al. DrugBank: a comprehensive resource for in silico drug discovery and exploration . Nucleic Acids Research 34 , D668 – D672 ( 2006 ). OpenUrl CrossRef PubMed Web of Science 58. ↵ Buniello , A. et al. Open Targets Platform: facilitating therapeutic hypotheses building in drug discovery . Nucleic Acids Research gka e1128 ( 2024 ) doi: 10.1093/nar/gkae1128 . OpenUrl CrossRef 59. ↵ Karhunen , V. , Bakker , M. K. , Ruigrok , Y. M. , Gill , D. & Larsson , S. C . Modifiable Risk Factors for Intracranial Aneurysm and Aneurysmal Subarachnoid Hemorrhage: A Mendelian Randomization Study . Journal of the American Heart Association 10 , e022277 ( 2021 ). OpenUrl CrossRef PubMed 60. ↵ Liu , H. et al. Genetically determined blood pressure, antihypertensive medications, and risk of intracranial aneurysms and aneurysmal subarachnoid hemorrhage: A Mendelian randomization study . Eur Stroke J 9 , 244 – 250 ( 2024 ). OpenUrl PubMed 61. ↵ Su , C.-Y. et al. Disentangling osteoarthritis-specific genetic effects from obesity to identify novel therapeutic targets . 2025.09.23.25336398 Preprint at doi: 10.1101/2025.09.23.25336398 ( 2025 ). OpenUrl Abstract / FREE Full Text 62. ↵ Lu , T. , Forgetta , V. , Greenwood , C. M. T. & Richards , J. B . Identifying Causes of Fracture Beyond Bone Mineral Density: Evidence From Human Genetics . Journal of Bone and Mineral Research 37 , 1592 – 1602 ( 2022 ). OpenUrl PubMed 63. ↵ Yin , S.-J. , Chou , C.-F. , Lai , C.-L. , Lee , S.-L. & Han , C.-L . Human class IV alcohol dehydrogenase: kinetic mechanism, functional roles and medical relevance . Chem Biol Interact 143–144, 219 – 227 ( 2003 ). 64. ↵ Kumar , S. , Sandell , L. L. , Trainor , P. A. , Koentgen , F. & Duester , G . Alcohol and aldehyde dehydrogenases: Retinoid metabolic effects in mouse knockout models . Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 1821 , 198 – 205 ( 2012 ). OpenUrl 65. ↵ Xiao , J. et al. All-trans retinoic acid attenuates the progression of Ang II-induced abdominal aortic aneurysms in ApoE−/−mice . J Cardiothorac Surg 15 , 160 ( 2020 ). 66. ↵ Lateef , H. , Stevens , M. J. & Varani , J . All- trans -Retinoic Acid Suppresses Matrix Metalloproteinase Activity and Increases Collagen Synthesis in Diabetic Human Skin in Organ Culture . The American Journal of Pathology 165 , 167 – 174 ( 2004 ). OpenUrl CrossRef PubMed 67. ↵ Axel , D. I. et al. All-trans retinoic acid regulates proliferation, migration, differentiation, and extracellular matrix turnover of human arterial smooth muscle cells . Cardiovasc Res 49 , 851 – 862 ( 2001 ). OpenUrl CrossRef PubMed 68. ↵ Goulden , B. D. et al. A high-avidity biosensor reveals plasma membrane PI(3,4)P2 is predominantly a class I PI3K signaling product . J Cell Biol 218 , 1066 – 1079 ( 2019 ). OpenUrl Abstract / FREE Full Text 69. ↵ Dowler , S. et al. Identification of pleckstrin-homology-domain-containing proteins with novel phosphoinositide-binding specificities . Biochem J 351 , 19 – 31 ( 2000 ). OpenUrl Abstract / FREE Full Text 70. ↵ Graupera , M. et al. Angiogenesis selectively requires the p110alpha isoform of PI3K to control endothelial cell migration . Nature 453 , 662 – 666 ( 2008 ). OpenUrl CrossRef PubMed Web of Science 71. ↵ Posor , Y. , Jang , W. & Haucke , V . Phosphoinositides as membrane organizers . Nat Rev Mol Cell Biol 23 , 797 – 816 ( 2022 ). OpenUrl CrossRef PubMed 72. ↵ Swaroop , A. , Branham , K. E. , Chen , W. & Abecasis , G . Genetic susceptibility to age-related macular degeneration: a paradigm for dissecting complex disease traits . Hum Mol Genet 16 , R174 – R182 ( 2007 ). OpenUrl CrossRef PubMed Web of Science 73. ↵ Goulden , B. D. et al. A high-avidity biosensor reveals plasma membrane PI(3,4)P2 is predominantly a class I PI3K signaling product . J Cell Biol 218 , 1066 – 1079 ( 2019 ). OpenUrl Abstract / FREE Full Text 74. ↵ High , F. A. et al. Endothelial expression of the Notch ligand Jagged1 is required for vascular smooth muscle development . Proceedings of the National Academy of Sciences 105 , 1955 – 1959 ( 2008 ). OpenUrl Abstract / FREE Full Text 75. ↵ Baeten , J. T. & Lilly , B . Notch Signaling in Vascular Smooth Muscle Cells . Adv Pharmacol 78 , 351 – 382 ( 2017 ). OpenUrl CrossRef PubMed 76. ↵ Ding , H. et al. Endothelial dysfunction in Type 2 diabetes correlates with deregulated expression of the tail-anchored membrane protein SLMAP . American Journal of Physiology-Heart and Circulatory Physiology 289 , H206 – H211 ( 2005 ). OpenUrl CrossRef PubMed 77. ↵ Mohamed Farhan , H. , Nassar , M. , Hassan Ahmed , M. , Abougabal , K. & Abd Elazim Taha , N. An association between the sarcolemmal membrane-associated protein gene and microvascular endothelial diabetic retinopathy in patients with type 2 diabetes mellitus: A preliminary case control study . Diabetes Metab Syndr 16 , 102653 ( 2022 ). View the discussion thread. Back to top Previous Next Posted November 13, 2025. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms Chen-Yang Su , Juliano Malizia , Masashi Hasebe , Thomas Zheng , Alejandro-Mejia Garcia , Hsuan Megan Tsao , Zhaohe Lin , Ta-Yu Yang , Fumihiko Matsuda , Patrick A. Dion , Vincent Mooser , Guy Rouleau , Guillaume Butler-Laporte , Tianyuan Lu , Satoshi Yoshiji , Sirui Zhou medRxiv 2025.11.11.25339992; doi: https://doi.org/10.1101/2025.11.11.25339992 Share This Article: Copy Citation Tools Multi-ancestry proteogenomic analysis identifies risk proteins for intracranial aneurysms Chen-Yang Su , Juliano Malizia , Masashi Hasebe , Thomas Zheng , Alejandro-Mejia Garcia , Hsuan Megan Tsao , Zhaohe Lin , Ta-Yu Yang , Fumihiko Matsuda , Patrick A. Dion , Vincent Mooser , Guy Rouleau , Guillaume Butler-Laporte , Tianyuan Lu , Satoshi Yoshiji , Sirui Zhou medRxiv 2025.11.11.25339992; doi: https://doi.org/10.1101/2025.11.11.25339992 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Genetic and Genomic Medicine Subject Areas All Articles Addiction Medicine (567) Allergy and Immunology (863) Anesthesia (297) Cardiovascular Medicine (4411) Dentistry and Oral Medicine (443) Dermatology (380) Emergency Medicine (606) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1505) Epidemiology (15205) Forensic Medicine (30) Gastroenterology (1119) Genetic and Genomic Medicine (6574) Geriatric Medicine (666) Health Economics (994) Health Informatics (4511) Health Policy (1365) Health Systems and Quality Improvement (1608) Hematology (537) HIV/AIDS (1263) Infectious Diseases (except HIV/AIDS) (15903) Intensive Care and Critical Care Medicine (1103) Medical Education (620) Medical Ethics (144) Nephrology (665) Neurology (6573) Nursing (345) Nutrition (998) Obstetrics and Gynecology (1139) Occupational and Environmental Health (954) Oncology (3319) Ophthalmology (967) Orthopedics (369) Otolaryngology (420) Pain Medicine (435) Palliative Medicine (129) Pathology (662) Pediatrics (1689) Pharmacology and Therapeutics (691) Primary Care Research (710) Psychiatry and Clinical Psychology (5421) Public and Global Health (9205) Radiology and Imaging (2191) Rehabilitation Medicine and Physical Therapy (1367) Respiratory Medicine (1191) Rheumatology (593) Sexual and Reproductive Health (709) Sports Medicine (529) Surgery (709) Toxicology (99) Transplantation (288) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe8dd33ddd20708',t:'MTc3OTI1NDEzMQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();