Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2025 Apr 23;20(4):e0320479. doi: 10.1371/journal.pone.0320479

Evidence to shared genetic correlation of ischemic stroke and intracerebral hemorrhage and cardiovascular related traits

Wei He 1,#, Jiajia Shi 2,#, Yiming Qian 1, Tao Fan 3, Xuehong Cai 4, Haochang Li 4, Peng Huang 4, Qin Shi 3,*
Editor: Francesco Sessa5
PMCID: PMC12017486  PMID: 40267100

Abstract

Background

Previous studies have demonstrated the genetic basis of stroke and also revealed their genetic correlation with some cardiovascular related diseases or traits at the entire genome, which, however, would not give the answer which regions may mainly account for the genetic overlap. This study aims to identify specific genetic loci that contribute to the shared genetic basis between ischemic stroke subtypes and common cardiovascular traits.

Methods

We used Local Analysis of [co]Variant Annotation (LAVA), a recent developed local genetic correlation method, to perform a system local genetic correlation analysis on GWAS summary data of two major subtypes of stroke, including any ischemic stroke (AIS) and intracerebral hemorrhage (ICH), and ten common cardiovascular related diseases or traits (CRTs). We further used colocalization analysis to explore potential shared causal genes in loci with significant local genetic correlation. In addition, we also performed Transcriptome-wide association (TWAS) analysis and fine-mapping for each phenotype to functionally annotate significant loci.

Results

LAVA analysis identified a total of 3 significant local genetic correlations (Bonferroni-adjusted P <  0.05) across 3 chromosomes between AIS and systolic blood pressure (SBP), AIS and hypertension (HT), and ICH and body mass index (BMI), among which locus 7.24 explicated to harbor a shared causal variant for AIS and SBP. TWIST1 in locus 7.24 was defined to be nominally associated with SBP, but not for AIS. Fine-mapping analysis also only identified TWIST1 a credible causal gene for BMI.

Conclusions

Our study revealed the local genetic correlations between two stroke subtypes and ten common CRTs. Gene-level analyses indicated that biological explanations underlying these identified local genetic correlations may existed elsewhere beyond a common pattern of genetic-gene expression regulation.

Introduction

Stroke is still a leading cause of both disability and death worldwide, and a major contributor to cognitive decline and dementia, especially in middle- and low-income countries [1,2]. The primary types of stroke are ischemic stroke (IS) and hemorrhagic stroke (e.g., intracerebral hemorrhage [ICH] and subarachnoid hemorrhage [SAH]) [3]. Considering the mechanism of stroke, emerging epidemiological studies have investigated the comorbidity between stroke and cardiovascular-related diseases or traits (CRTs), such as blood cholesterol, body mass index (BMI), atrial fibrillation (AF), and hypertension (HT) [4–6]. Beyond phenotypic comorbidity, the genetic correlation, a crucial aspect of comorbidity, is useful in the mechanism of stroke and other complex diseases [7]. Genome-wide association studies (GWAS), defining numbers of risk loci, provided an opportunity to detect genetic architecture of both stroke and CRTs [8–11]. Given the shared pathological features of stroke and CRTs affecting blood vessels [12], understanding these associations is crucial, as it may reveal shared pathophysiological mechanisms that contribute to both stroke and cardiovascular diseases, thereby informing preventive strategies [13].

Although previous studies have estimated genetic correlations (rg) between stroke and CRTs, they primarily employed global rg approaches [14–19]. Local rg in the absence of any global correlation may be undetected. Recently developed methods for assessing local genetic correlations allow for a more nuanced understanding of the shared genetic architecture among complex diseases [19–22]. On the other hand, the correlated regions provide more evidence for the common genes defined by transcriptome-wide association (TWAS). We aim to explore the specific genetic loci shared between two stroke subtypes (AIS and ICH) and seven CRTs, thereby advancing our understanding of their comorbidity.

Here, using accessible GWAS summary statistics, we performed a systemical analysis for the comorbidities between two stroke subtypes, including IS and ICH, and ten CRTs. First, we employed Local Analysis of [co]Variant Annotation (LAVA) [19], to perform a system local genetic correlation analysis. Different to the genome-wide genetic correlation, LAVA employs singular value decomposition to estimate the variance of SNPs after partitioning the genome into semi-independent LD blocks [19]. It enables us to pinpoint shared genetic variants and potential functional genes, facilitating a deeper understanding of the biological mechanisms underlying complex diseases. Secondly, we used colocalization analysis to explore whether the identified loci with significant local genetic correlation contained shared causal genes for both phenotypes. Finally, we performed TWAS analysis and fine-mapping for each phenotype to functionally annotate significant loci.

Methods and materials

GWAS summary statistics

We downloaded GWAS data two major subtypes of stroke, including AIS [10] and ICH [23], and ten CRTs, including total cholesterol (TC) [24], HDL [24], low density lipoprotein cholesterol (LDL) [24], logarithm of triglycerides (logTG) [24], BMI [25]; systolic blood pressure (SBP) [26], diastolic blood pressure (DBP) [26], AF [27], HT [28], and familial combined hyperlipidemia (FHL) [29]. Twelve GWASs adopted including from 3,026 to 1,320,016 European individuals. We summarized the detailed demographic information, such as the ratio of sex and heritability, available in the cited studies (S1 Table) [10,23–29]. Specifically, we used the summary statistics for eQTL from GTEx V8, which were available without any application. All cohort data and GWAS resources were approved by relevant ethics committees, and written informed consent was obtained from all participants. We further excluded SNPs with MAF <  0.001, and 5.170 to 11.993 million SNPs were retained for subsequent analysis. A summary of these GWAS summary data is shown in S1 Table. Our data come from the public database where participants have been ethically approved.

Global genetic correlation analysis

We used linkage disequilibrium score regression (LDSC, v 1.0.1) to estimate (i) the SNP-based heritability (h2SNP) for each phenotype, (ii) the global genetic correlation (rg) between each phenotype pair, and (iii) the sample overlap [30,31]. We also used 1000 Genomes Project Phase 3 European population as reference panel [32]. Significant SNP-based heritability was defined as with z score >  2, while significant global genetic correlation was defined as with Bonferroni-adjusted P value <  0.05.

Local genetic correlation analysis

We used LAVA package (v 0.1.0) to perform local genetic correlation analysis between phenotype pairs [19]. In brief, given the summary statistic of two phenotypes and a locus defined by genomic coordinates or a list of SNPs, LAVA can estimate the standard bivariate local rg between the two phenotypes at the defined locus, while accounting for known or estimated sample overlap [19]. In specific, we leveraged the sample overlap estimated by LDSC, and used the predefined loci file by LAVA, which contained 2495 semi-independent LD blocks with a minimum block size of 2500 base pairs (https://github.com/josefin-werme/LAVA/tree/main/support_data) [19,33]. We used run.univ.bivar function to perform univariate and bivariate test sequentially. P value threshold for the univariate test was set to 0.05/2495 to account for the number of loci tested, and bivariate test was then performed only for the phenotypes that reach the desired univariate significance threshold. Significant bivariate loci were defined as those with Bonferroni-adjusted P value <  0.05. Nominal bivariate loci were defined as those with unadjusted P value <  0.05, which was used to the following analytic procedures.

Colocalization analysis

We used coloc package (v 5.2.2) to perform Approximate Bayes Factor (ABF) colocalization analysis [34]. In brief, for each nominal bivariate locus detected by LAVA, we used the coloc.abf function to assessed the posterior probability that the two phenotypes share same (PPH4) or different (PPH3) causal variants in this locus [34]. We set the prior probabilities as recommended: (1) 1E-4 for a variant associated with either trait; (2) 1E-5 for a variant associated with both traits; (3) a posterior probability larger than 0.5 was considered as evidence for colocalization.

TWAS

We performed tissue-specific TWAS analysis using functional summary-based expression imputation algorithm (FUSION) to identify genetic-predicted gene expression associated with each phenotype [35]. We used the tissue-specific predictive models pre-computed on GTEx V8 EUR individuals by the Mancuso lab as TWAS weights. In specific, we downloaded those on 13 different tissues, including eight brain tissues, two heart tissues and one on whole blood, from http://gusevlab.org/projects/fusion/#gtex-v8-multi-tissue-expression. Each weight file contains models for top eQTL, elastic net [36], Sum of Single Effects (SuSiE) [37], and least absolute shrinkage and selection operator (LASSO) [38,39]. For each genetic-predicted gene expression-phenotype pair, FUSION then imputed GWAS Z-scores using the IMPG algorithm; estimated the association statistic, and reported the results of the model with best performance. For each phenotype, significant genes were defined as those with Bonferroni-adjusted P value <  0.05.

Functional analysis of bivariate loci

We used Fine-mapping Of CaUsal gene Sets (FOCUS) to prioritize causal genes within each significant bivariate loci [40]. We downloaded the recommended multi-tissue gene expression weights (https://github.com/bogdanlab/focus/wiki), which combined GTEx V7 weights from PrediXcan with METSIM, NTR, YFS, and CMC weights from FUSION together. FOCUS then calculated the posterior inclusion probability (PIP) for each gene and output a credible set of genes to explain observed genomic risk. Genes with PIP >  0.5 were defined as causal ones.

Results

Global and local genetic correlations

We first explored the genetic correlation among the two stroke subtypes and ten CRTs. The results showed that all phenotypes exhibited significant SNP-based heritability, among which the heritability of AIS and ICH were 0.0135 (Z score =  9.6429) and 0.3959 (Z score =  2.6570), respectively (S1 Table). Seven CRTs, including logTG, BMI, SBP, DBP, AF, HT, and FHL, exhibited significantly positive global genetic correlation with AIS, while no CRT showed significant correlation with ICH (S2 Table and Fig 1). We further detected three significant local genetic correlations, which are between AIS and SBP (r =  0.7480, P value =  1.62 × 10-8, Padj =  1.42 × 10-4), AIS and HT (r =  0.8099, P value =  2.66 × 10-6, Padj =  2.35 × 10-2), and ICH and BMI (r =  0.8067, P value =  1.75 × 10-6, Padj =  1.54 × 10-2), respectively (S3 Table and Fig 1). In addition, 124 local genetic correlations showed only nominal significance (unadjusted P value <  0.05), of which 40 and 22 are on BMI and DBP, respectively (S3 Table and Fig 1). Of note, 11 of the nominally significant local genetic correlations showed discordant direction with those of corresponding global correlation that showed statistical significance (S2 and S3 Tables).

Fig 1. LDSC and LAVA correlations across 12 diseases or traits.

Fig 1

Open in a new tab

The lower triangle displays the global (LDSC) genetic correlations for trait pairs while the upper triangle displays the number of significant bivariate local (LAVA) genetic correlations for each trait pair. Abbreviations: AIS: any ischemic stroke; ICH: intracerebral hemorrhage; TC: total cholesterol; HDL: high density lipoprotein cholesterol; LDL: low density lipoprotein cholesterol; logTG: logarithm of triglycerides; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; AF: atrial fibrillation; HT: hypertension; FHL: familial combined hyperlipidemia.

Colocalization analysis

We then performed colocalization to evaluate the posterior probabilities that at least nominally significant bivariate loci comprises independent (H3) or shared casual variant (H4) for two corresponding phenotypes. The results showed that one of the three significant bivariate loci contained a shared causal variant for AIS and SBP (locus 7.24 with PPH4 =  0.999), and two nominal bivariate loci also contained a shared causal variant, among which one is between AIS and HT (locus 7.24 with PPH4 =  0.996), and another is between AIS and DBP (locus 13.86 with PPH4 =  0.913) (S4 Table and Fig 2). In addition, six nominally significant local genetic correlations indicated potential independent causal variant (with PPH3 >  0.5), among which five were on AIS and the highest is between AIS and HDL (locus 7.24 with PPH3 =  1.000) (S4 Table and Table 1).

Fig 2. Colocalization analysis of significant bivariate loci.

Fig 2

Open in a new tab

Abbreviations: AIS: any ischemic stroke; ICH: intracerebral hemorrhage; TC: total cholesterol; HDL: high density lipoprotein cholesterol; LDL: low density lipoprotein cholesterol; logTG: logarithm of triglycerides; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; AF: atrial fibrillation; HT: hypertension; FHL: familial combined hyperlipidemia; PP, posterior probability.

Table 1. The summary of pleiotropic loci (PPH3/PPH4 >  0.5).

Locus Phenotype1 Phenotype2 LAVA.P PP.H3 PP.H4 TWAS top gene of Phenotype1 Top TWAS.P of Phenotype1 PIP of Phenotype1 TWAS top gene of Phenotype2 Top TWAS.P of Phenotype2 PIP of Phenotype2
7.24 SBP AIS 1.62E-08 0.001 0.999 TWIST1 0.031 0.149 TWIST1 0.003 0.136
7.24 HT AIS 1.35E-05 0.002 0.996 SNX13 0.261 0.014 TWIST1 0.003 0.136
13.86 DBP AIS 0.009 0.029 0.913 LINC00346 0.048 0.022 ANKRD10 0.017 0.005
7.24 HDL AIS 4.66E-04 1.000 0 SNX13 2.56E-14 0.035 TWIST1 0.003 0.136
9.29 DBP AIS 0.048 0.992 0 MTAP 1.73E-10 0.024 CDKN2A 9.12E-06 0.015
12.55 LDL AIS 0.015 0.844 0.001 PIP4K2C 2.58E-07 0.072 TAC3 0.015 0.012
12.55 TC AIS 0.029 0.842 0.004 TAC3 2.72E-07 0.006 TAC3 0.015 0.012
12.55 SBP AIS 0.003 0.737 0.125 ARHGEF25 2.64E-09 0.047 TAC3 0.015 0.012
6.158 HDL ICH 0.11 0.568 0.025 SLC22A3 2.19E-15 0.086 RP11-288H12.3 0.002 0.001
Open in a new tab

TWAS

In tissue-specific TWAS analysis for each phenotype, we mainly focused on those within significant bivariate loci. PRKCE in locus 2.51, a significant bivariate locus for AIS and HT, showed positive association with HT in whole blood (Z score of TWAS =  5.83, and Bonferroni-adjusted P value =  1.717E-10), but not with AIS. In addition, TWIST in locus 7.24, a significant bivariate locus for AIS and SBP, showed nominally negative association with SBP in brain spinal cord cervical c-1 (Z score =  -2.17, and unadjusted P value =  2.99E-02), but not with AIS (S5 and S6 Tables and Fig 3). However, we also detected significant genes in different tissues for nominally significant bivariate loci. For example, SNX13 in locus 7.24 for AIS and HDL showed negative association with HDL in whole blood (Z score of TWAS =  -7.62, and Bonferroni-adjusted P value =  8.19E-13) (Table 1).

Fig 3. Heatmap indicating concordance of LAVA and TWAS results.

Fig 3

Open in a new tab

The upper heatmap denotes TWAS FUSION z scores for genes within significant LAVA bivariate loci, with “*” indicating significant genes in TWAS analysis (adjusted P value <  0.05), and “·” indicating nominally significant genes in TWAS analysis (unadjusted P value <  0.05). The lower annotation grid denotes the significance of bivariate phenotype pairs, with black indicating significant pair. Abbreviations: AIS: any ischemic stroke; ICH: intracerebral hemorrhage; TC: total cholesterol; HDL: high density lipoprotein cholesterol; LDL: low density lipoprotein cholesterol; logTG: logarithm of triglycerides; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; AF: atrial fibrillation; HT: hypertension; FHL: familial combined hyperlipidemia.

Concordance of genetic and gene-based correlations

All three significant bivariate loci identified by LAVA showed relative concordant direction with the global genetic correlations of the corresponding phenotype pair identified by LDSC (S7 Table). We also evaluated the Pearson’s correlation of TWAS statistics Z of genes within a given bivariate locus for the corresponding phenotype pair. As only one and two genes located in locus 2.51 and 7.24, we only assessed the concordance of global and local genetic correlations with the defined TWAS correlation for locus 10.11. However, different from the positive correlation defined by LAVA and LDSC, we did not find any evidence of correlation (S7 Table).

Gene prioritization

Finally, we performed fine-mapping to prioritize gene-phenotype associations within each bivariate locus. Unexpectedly, only TWIST1 in locus 7.24 showed to be a credible causal gene of BMI in nerve tibial (PIP =  0.566), although it was also the gene with the highest PIP for both SBP (PIP =  0.149) and AIS (PIP =  0.136) in artery tibial (S8 Table).

Discussion

In this study, we performed local genetic correlation analysis on two stroke subtypes and ten CRTs. We identified a total of three significant bivariate loci on AIS or ICH. Of note, we also found 11 nominal bivariate loci showed discordant direction with those of corresponding global correlation, indicating that a potential heterogeneity across internal regions existed even between phenotype pairs that exhibited significantly global genetic correlation, and underlined the importance of local genetic correlation analysis in exploring shared genetic basis. Interestingly, our findings were consistent with the previous studies. For example, Laura et al. found that the genetic architecture of stroke risk is correlated with that of HT [16]. The primary reason is that they overlap in multiple biological mechanisms, including vascular health, coagulation function, inflammatory response, and metabolic disorders. Different stroke subtypes may act through these common genetic risk factors, but they may also lead to different types of strokes due to distinct pathological processes. Additionally, the interaction between genetics and the environment, as well as the effects of epigenetics, may produce similar genetic effects across various subtypes. Therefore, the genetic susceptibility to stroke is multifaceted, and there is a high degree of genetic overlap between different subtypes.

The locus 7.24 was defined to harbor a significant local genetic correlation for AIS and SBP, and two nominal local genetic correlations for AIS and HDL and AIS and HT. Using colocalization analysis, we found two phenotype pairs (i.e., AIS and SBP and AIS and HT) had a shared causal variant in the locus 7.24, while AIS and HDL may have an independent causal variant in this locus. These results suggested this locus an important contributor to the shared basic of AIS and CRTs. SNX13 and TWIST1 are two protein coding genes known to be within locus 7.24 or overlap the boundary of this locus. SNX13 encodes a protein containing phox-homology domain and the regulator of G protein signaling (RGS) domain, belonging to both the sorting nexin (SNX) and RGS protein families [41]. SNX13 has been demonstrated to play a potential role in inter-organelle communication and lipid metabolism, and may be involved in the cardiac performance and pancreatic islet function [41–44]. TWIST1 encodes Twist family BHLH transcription factor 1, and was reported as hypermethylated and overexpressed in multiple human cancers, including lung cancer, prostate cancer, and breast cancer [45–47]. Recent studies also found the contribution of TWIST1 in vascular diseases like pulmonary hypertension, which may be due to promoting the proliferation of smooth muscle cell [48–50]. In our TWAS analysis, although the genetic predicted expression of SNX13 and TWIST1 were found to be significantly associated with HDL and nominally associated with SBP, respectively, we did not identify their association with AIS. Our fine-mapping analysis also only identified TWIST1 a credible causal gene for BMI. These results may not support common genetic-gene expression regulation patterns underlying the shared genetic basis among these stroke subtypes and CRTs.

To our best knowledge, our study is the first to take into consideration the local genetic correlation between stroke and CRTs, enabling it to provide some additional clues beyond global genetic correlation analysis for understanding the shared genetic basis between stoke and CRTs. To maximize statistical power and representativeness, we included GWAS summary data with nearly the largest sample size publicly available. In addition, for identified loci with even only nominal significance, we employed a series of validation tools, including colocalization, TWAS, and fine-mapping, to further explore the potential key genetic variants or genes that may drive the local genetical correlation. But several limitations of our study warrant consideration. First, our analysis is based on GWAS data predominantly from individuals of European ancestry. It may restrict the generalizability of our findings to other ethnic groups and bring statistical bias in the observed genetic correlations [14,51]. Second, because of the low prevalence and distinct pathophysiological mechanisms, we primarily focused on AIS and ICH, rather than SAH. The low case number in the GWAS summary statistics make it difficult to analyze the genetic associations [52]. Third, due to the presence of different subtypes of stroke, there may be heterogeneity within our study population, which could lead to incomplete results. Therefore, future research should consider these differences in the design, data analysis, and interpretation of results, in order to draw more accurate, comprehensive, and meaningful conclusions [53,54]. Finally, with our knowledge, cardiovascular diseases might be more closely related to ICH. But SAH is commonly seen in conditions such as aneurysms and arteriovenous malformations. Third, the extensive multiple testing we made may also decrease the statistical power in local genetic correlation. To alleviate this issue, we also pay considerable attention to those nominally significant loci in our analysis. The integration of genetic insights with clinical data could enhance our understanding of stroke mechanisms and guide personalized interventions.

In summary, we explored the shared genetic basis of two stroke subtypes and ten CRTs in a series of semi-independent genomic loci via local genetic correlation analysis. We found a strong stroke-CRT genetic correlation at a total three loci, and also found a potential heterogeneity across internal regions between phenotype pairs showing significant global correlation. Gene-level analyses did not identify potential pleiotropic genes, indicating that the biological explanations underlying these identified local genetic correlations may exist elsewhere beyond a common pattern of genetic-gene expression regulation, and further exploration is needed.

Supporting information

S1 Table. Phenotype and sample information.

(XLSX)

pone.0320479.s001.xlsx (11.6KB, xlsx)
S2 Table. LDSC genetic correlations across 12 diseases or traits.

(XLSX)

pone.0320479.s002.xlsx (20.8KB, xlsx)
S3 Table. LAVA local genetic correlations across 12 diseases or traits.

(XLSX)

pone.0320479.s003.xlsx (216.7KB, xlsx)
S4 Table. COLOC results for loci with a nominally significant local genetic correlation.

(XLSX)

pone.0320479.s004.xlsx (23.6KB, xlsx)
S5 Table. TWAS FUSION results for 12 phenotypes.

(XLSX)

pone.0320479.s005.xlsx (55.2KB, xlsx)
S6 Table. TWAS FUSION Z scores for genes within the bivariate loci.

(XLSX)

pone.0320479.s006.xlsx (12.4KB, xlsx)
S7 Table. Comparison of LDSC, LAVA, and TWAS correlations.

(XLSX)

pone.0320479.s007.xlsx (10.6KB, xlsx)
S8 Table. FOCUS fine-mapping results.

(XLSX)

pone.0320479.s008.xlsx (10.6KB, xlsx)

Data Availability

Pre-computed 13 tissue-specific TWAS weights on GTEx V8 EUR individuals were downloaded from http://gusevlab.org/projects/fusion/#gtex-v8-multi-tissue-expression. Multi-tissue gene expression weights for FOCUS were downloaded from https://github.com/bogdanlab/focus/wiki.

Funding Statement

The author(s) received no specific funding for this work.

References

  • 1.Feigin V, Brainin M, Norrving B, Martins S, Sacco R, Hacke W, et al. Global stroke fact sheet 2022. Inter J Stroke. 2022;17(1):18–29. [DOI] [PubMed] [Google Scholar]
  • 2.Capizzi A, Woo J, Verduzco-Gutierrez M. Traumatic brain injury: an overview of epidemiology, pathophysiology, and medical management. Med Clin North Am. 2020;104(2):213–38. doi: 10.1016/j.mcna.2019.11.001 [DOI] [PubMed] [Google Scholar]
  • 3.Debette S, Markus HS. Stroke genetics: discovery, insight into mechanisms, and clinical perspectives. Circ Res. 2022;130(8):1095–111. doi: 10.1161/CIRCRESAHA.122.319950 [DOI] [PubMed] [Google Scholar]
  • 4.Sun L, Clarke R, Bennett D, Guo Y, Walters RG, Hill M, et al. Causal associations of blood lipids with risk of ischemic stroke and intracerebral hemorrhage in Chinese adults. Nat Med. 2019;25(4):569–74. doi: 10.1038/s41591-019-0366-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Yuan S, Huang X, Ma W, Yang R, Xu F, Han D, et al. Associations of HDL-C/LDL-C with myocardial infarction, all-cause mortality, haemorrhagic stroke and ischaemic stroke: a longitudinal study based on 384 093 participants from the UK Biobank. Stroke Vasc Neurol. 2023;8(2):119–26. doi: 10.1136/svn-2022-001668 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Vargas-González J-C, Hachinski V. Insidious cerebrovascular disease-the uncool iceberg. JAMA Neurol. 2020;77(2):155–6. doi: 10.1001/jamaneurol.2019.3933 [DOI] [PubMed] [Google Scholar]
  • 7.Cai H, Cai B, Liu Z, Wu W, Chen D, Fang L, et al. Genetic correlations and causal inferences in ischemic stroke. J Neurol. 2020;267(7):1980–90. doi: 10.1007/s00415-020-09786-4 [DOI] [PubMed] [Google Scholar]
  • 8.Bakker MK, van der Spek RAA, van Rheenen W, Morel S, Bourcier R, Hostettler IC, et al. Genome-wide association study of intracranial aneurysms identifies 17 risk loci and genetic overlap with clinical risk factors. Nat Genet. 2020;52(12):1303–13. doi: 10.1038/s41588-020-00725-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mishra A, Malik R, Hachiya T, Jurgenson T, Namba S, Posner D, et al. Stroke genetics informs drug discovery and risk prediction across ancestries. Nature. 2022;611(7934):115–23. doi: insert_doi_here_if_available [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Malik R, Chauhan G, Traylor M, Sargurupremraj M, Okada Y, Mishra A, et al. Multiancestry genome-wide association study of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat Genet. 2018;50(4):524–37. doi: 10.1038/s41588-018-0058-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kals M, Kunzmann K, Parodi L, Radmanesh F, Wilson L, Izzy S, et al. A genome-wide association study of outcome from traumatic brain injury. EBioMedicine. 2022;77:103933. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Andjelkovic AV, Keep RF, Wang MM. Molecular mechanisms of cerebrovascular diseases. Int J Mol Sci. 2022;23(13). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cao C, Tian M, Li Z, Zhu W, Huang P, Yang S. GWAShug: a comprehensive platform for decoding the shared genetic basis between complex traits based on summary statistics. Nucleic Acids Res. 2025;53(D1):D1006–15. doi: 10.1093/nar/gkae873 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Cao C, Zhang S, Wang J, Tian M, Ji X, Huang D, et al. PGS-Depot: a comprehensive resource for polygenic scores constructed by summary statistics based methods. Nucleic Acids Res. 2024;52(D1):D963–D71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Yang S, Zhou X. Accurate and scalable construction of polygenic scores in large biobank data sets. Am J Hum Genet. 2020;106(5):679–93. doi: 10.1016/j.ajhg.2020.03.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Ibanez L, Heitsch L, Dube U, Farias FHG, Budde J, Bergmann K, et al. Overlap in the genetic architecture of stroke risk, early neurological changes, and cardiovascular risk factors. Stroke. 2019;50(6):1339–45. doi: 10.1161/STROKEAHA.118.023097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Harshfield EL, Georgakis MK, Malik R, Dichgans M, Markus HS. Modifiable lifestyle factors and risk of stroke: a mendelian randomization analysis. Stroke. 2021;52(3):931–6. doi: 10.1161/STROKEAHA.120.031710 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Pulit SL, Weng L-C, McArdle PF, Trinquart L, Choi SH, Mitchell BD, et al. Atrial fibrillation genetic risk differentiates cardioembolic stroke from other stroke subtypes. Neurol Genet. 2018;4(6):e293. doi: 10.1212/NXG.0000000000000293 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Werme J, van der Sluis S, Posthuma D, de Leeuw CA. An integrated framework for local genetic correlation analysis. Nat Genet. 2022;54(3):274–82. doi: 10.1038/s41588-022-01017-y [DOI] [PubMed] [Google Scholar]
  • 20.Zhang Y, Lu Q, Ye Y, Huang K, Liu W, Wu Y, et al. SUPERGNOVA: local genetic correlation analysis reveals heterogeneous etiologic sharing of complex traits. Genome Biol. 2021;22(1):262. doi: 10.1186/s13059-021-02478-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Guo H, Li JJ, Lu Q, Hou L. Detecting local genetic correlations with scan statistics. Nat Commun. 2021;12(1):2033. doi: 10.1038/s41467-021-22334-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Shi H, Mancuso N, Spendlove S, Pasaniuc B. Local genetic correlation gives insights into the shared genetic architecture of complex traits. Am J Hum Genet. 2017;101(5):737–51. doi: 10.1016/j.ajhg.2017.09.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Woo D, Falcone GJ, Devan WJ, Brown WM, Biffi A, Howard TD, et al. Meta-analysis of genome-wide association studies identifies 1q22 as a susceptibility locus for intracerebral hemorrhage. Am J Hum Genet. 2014;94(4):511–21. doi: 10.1016/j.ajhg.2014.02.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Graham SE, Clarke SL, Wu K-HH, Kanoni S, Zajac GJM, Ramdas S, et al. The power of genetic diversity in genome-wide association studies of lipids. Nature. 2021;600(7890):675–9. doi: 10.1038/s41586-021-04064-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Pulit SL, Stoneman C, Morris AP, Wood AR, Glastonbury CA, Tyrrell J, et al. Meta-analysis of genome-wide association studies for body fat distribution in 694 649 individuals of European ancestry. Hum Mol Genet. 2019;28(1):166–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Evangelou E, Warren HR, Mosen-Ansorena D, Mifsud B, Pazoki R, Gao H, et al. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nat Genet. 2018;50(10):1412–25. doi: 10.1038/s41588-018-0205-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Nielsen JB, Thorolfsdottir RB, Fritsche LG, Zhou W, Skov MW, Graham SE, et al. Biobank-driven genomic discovery yields new insight into atrial fibrillation biology. Nat Genet. 2018;50(9):1234–9. doi: 10.1038/s41588-018-0171-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zhu Z, Wang X, Li X, Lin Y, Shen S, Liu C-L, et al. Genetic overlap of chronic obstructive pulmonary disease and cardiovascular disease-related traits: a large-scale genome-wide cross-trait analysis. Respir Res. 2019;20(1):64. doi: 10.1186/s12931-019-1036-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Trinder M, Vikulova D, Pimstone S, Mancini GBJ, Brunham LR. Polygenic architecture and cardiovascular risk of familial combined hyperlipidemia. Atherosclerosis. 2022;340:35–43. doi: 10.1016/j.atherosclerosis.2021.11.032 [DOI] [PubMed] [Google Scholar]
  • 30.Bulik-Sullivan BK, Loh P-R, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics Consortium, et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet. 2015;47(3):291–5. doi: 10.1038/ng.3211 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Gerring ZF, Thorp JG, Gamazon ER, Derks EM. A local genetic correlation analysis provides biological insights into the shared genetic architecture of psychiatric and substance use phenotypes. Biol Psychiatry. 2022;92(7):583–91. doi: 10.1016/j.biopsych.2022.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Auton A, Abecasis G, Altshuler D, Durbin R, Bentley D, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Bulik-Sullivan B, Finucane HK, Anttila V, Gusev A, Day FR, Loh P-R, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47(11):1236–41. doi: 10.1038/ng.3406 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Giambartolomei C, Vukcevic D, Schadt EE, Franke L, Hingorani AD, Wallace C, et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10(5):e1004383. doi: 10.1371/journal.pgen.1004383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Gusev A, Ko A, Shi H, Bhatia G, Chung W, Penninx BWJH, et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet. 2016;48(3):245–52. doi: 10.1038/ng.3506 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Zou H, Hastie T. Regularization and variable selection via the elastic net. J R Stat Soc Series B Stat Method. 2005;67(2):301–20. doi: 10.1111/j.1467-9868.2005.00503.x [DOI] [Google Scholar]
  • 37.Wang G, Sarkar A, Carbonetto P, Stephens M. A simple new approach to variable selection in regression, with application to genetic fine mapping. J R Stat Soc Series B Stat Methodol. 2020;82(5):1273–300. doi: 10.1111/rssb.12388 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tibshirani R. Regression shrinkage and selection via the lasso. J R Stati Soc Series B Stat Method. 1996;58(1):267–88. doi: 10.1111/j.2517-6161.1996.tb02080.x [DOI] [Google Scholar]
  • 39.Veturi Y, Ritchie M. How powerful are summary-based methods for identifying expression-trait associations under different genetic architectures? Pac Symp Biocomput. 2018;23:228–39. [PMC free article] [PubMed] [Google Scholar]
  • 40.Mancuso N, Freund MK, Johnson R, Shi H, Kichaev G, Gusev A, et al. Probabilistic fine-mapping of transcriptome-wide association studies. Nat Genet. 2019;51(4):675–82. doi: 10.1038/s41588-019-0367-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Li J, Li C, Zhang D, Shi D, Qi M, Feng J, et al. SNX13 reduction mediates heart failure through degradative sorting of apoptosis repressor with caspase recruitment domain. Nat Commun. 2014;5:5177. doi: 10.1038/ncomms6177 [DOI] [PubMed] [Google Scholar]
  • 42.Lu A, Hsieh F, Sharma B, Vaughn S, Enrich C, Pfeffer S. CRISPR screens for lipid regulators reveal a role for ER-bound SNX13 in lysosomal cholesterol export. J Cell Biol. 2021;221(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Henne WM, Zhu L, Balogi Z, Stefan C, Pleiss JA, Emr SD. Mdm1/Snx13 is a novel ER-endolysosomal interorganelle tethering protein. J Cell Biol. 2015;210(4):541–51. doi: 10.1083/jcb.201503088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Pedersen H, Gudmundsdottir V, Brunak S. Pancreatic islet protein complexes and their dysregulation in type 2 diabetes. Front Genet. 2017;8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Xu Y, Qin L, Sun T, Wu H, He T, Yang Z, et al. Twist1 promotes breast cancer invasion and metastasis by silencing Foxa1 expression. Oncogene. 2017;36(8):1157–66. doi: 10.1038/onc.2016.286 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Pan J, Fang S, Tian H, Zhou C, Zhao X, Tian H, et al. lncRNA JPX/miR-33a-5p/Twist1 axis regulates tumorigenesis and metastasis of lung cancer by activating Wnt/β-catenin signaling. Mol Cancer. 2020;19(1):9. doi: 10.1186/s12943-020-1133-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Lee Y, Yoon J, Ko D, Yu M, Lee S, Kim S. TMPRSS4 promotes cancer stem-like properties in prostate cancer cells through upregulation of SOX2 by SLUG and TWIST1. J Exp Clin Cancer Res. 2021;40(1):372. doi: 10.1186/s13046-021-02147-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Nurnberg ST, Guerraty MA, Wirka RC, Rao HS, Pjanic M, Norton S, et al. Genomic profiling of human vascular cells identifies TWIST1 as a causal gene for common vascular diseases. PLoS Genet. 2020;16(1):e1008538. doi: 10.1371/journal.pgen.1008538 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Fan Y, Gu X, Zhang J, Sinn K, Klepetko W, Wu N, et al. TWIST1 drives smooth muscle cell proliferation in pulmonary hypertension via loss of GATA-6 and BMPR2. Am J Respir Crit Care Med. 2020;202(9):1283–96. doi: 10.1164/rccm.201909-1884OC [DOI] [PubMed] [Google Scholar]
  • 50.Kyi P, Hendee K, Hunyenyiwa T, Matus K, Mammoto T, Mammoto A. Endothelial senescence mediates hypoxia-induced vascular remodeling by modulating PDGFB expression. Front Med. 2022;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Yang S, Zhou X. PGS-server: accuracy, robustness and transferability of polygenic score methods for biobank scale studies. Brief Bioinform. 2022;23(2). [DOI] [PubMed] [Google Scholar]
  • 52.Backman JD, Li AH, Marcketta A, Sun D, Mbatchou J, Kessler MD, et al. Exome sequencing and analysis of 454,787 UK Biobank participants. Nature. 2021;599(7886):628–34. doi: 10.1038/s41586-021-04103-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Yang J, Zhang B, Hu C, Jiang X, Shui P, Huang J, et al. Identification of clinical subphenotypes of sepsis after laparoscopic surgery. Laparosc Endosc Robot Surg. 2024;7(1):16–18. [Google Scholar]
  • 54.Yang S, Zhou X. SRT-Server: powering the analysis of spatial transcriptomic data. Genome Med. 2024;16(1):18. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Table. Phenotype and sample information.

(XLSX)

pone.0320479.s001.xlsx (11.6KB, xlsx)
S2 Table. LDSC genetic correlations across 12 diseases or traits.

(XLSX)

pone.0320479.s002.xlsx (20.8KB, xlsx)
S3 Table. LAVA local genetic correlations across 12 diseases or traits.

(XLSX)

pone.0320479.s003.xlsx (216.7KB, xlsx)
S4 Table. COLOC results for loci with a nominally significant local genetic correlation.

(XLSX)

pone.0320479.s004.xlsx (23.6KB, xlsx)
S5 Table. TWAS FUSION results for 12 phenotypes.

(XLSX)

pone.0320479.s005.xlsx (55.2KB, xlsx)
S6 Table. TWAS FUSION Z scores for genes within the bivariate loci.

(XLSX)

pone.0320479.s006.xlsx (12.4KB, xlsx)
S7 Table. Comparison of LDSC, LAVA, and TWAS correlations.

(XLSX)

pone.0320479.s007.xlsx (10.6KB, xlsx)
S8 Table. FOCUS fine-mapping results.

(XLSX)

pone.0320479.s008.xlsx (10.6KB, xlsx)

Data Availability Statement

Pre-computed 13 tissue-specific TWAS weights on GTEx V8 EUR individuals were downloaded from http://gusevlab.org/projects/fusion/#gtex-v8-multi-tissue-expression. Multi-tissue gene expression weights for FOCUS were downloaded from https://github.com/bogdanlab/focus/wiki.

Articles from PLOS One are provided here courtesy of PLOS

RESOURCES