Abstract
Fine-mapping in genome-wide association studies attempts to identify causal SNPs from a set of candidate SNPs in a local genomic region of interest and is commonly performed in one genetic ancestry at a time. Here, we present multi-ancestry sum of the single effects model (MESuSiE), a probabilistic multi-ancestry fine-mapping method, to improve the accuracy and resolution of fine-mapping by leveraging association information across ancestries. MESuSiE uses summary statistics as input, accounts for the diverse linkage disequilibrium pattern observed in different ancestries, explicitly models both shared and ancestry-specific causal SNPs, and relies on a variational inference algorithm for scalable computation. We evaluated the performance of MESuSiE through comprehensive simulations and multi-ancestry fine-mapping of four lipid traits with both European and African samples. In the real data, MESuSiE improves fine-mapping resolution by 19.0% to 72.0% compared to existing approaches, is an order of magnitude faster, and captures and categorizes shared and ancestry-specific causal signals with enhanced functional enrichment.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 12 print issues and online access
$209.00 per year
only $17.42 per issue
Buy this article
- Purchase on Springer Link
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
The individual-level genotype data UKBB are available at http://www.ukbiobank.ac.uk. GWAS summary statistics for the four lipid traits and MCHC in UKBB are available at http://www.nealelab.is/uk-biobank. GWAS summary statistics for the four lipid traits from GLGC are available at http://csg.sph.umich.edu/willer/public/glgc-lipids2021/. GWAS summary statistics of MCHC in Biobank Japan are available at https://pheweb.jp/downloads. GWAS summary statistics of SCZ are available at https://figshare.com/articles/dataset/scz2019asi/19193084. The GTEx independent eQTL v8 data are publicly available at https://gtexportal.org/home/datasets. Baseline annotation data are available at https://alkesgroup.broadinstitute.org/LDSCORE/baselineLD_v2.1_annots/. 1000 Genomes data (phase 3) are available at https://ftp.1000genomes.ebi. ac.uk/vol1/ftp/release/20130502/. The summary statistics data and the fine-mapping results presented in the study are also available at Zenodo55 https://doi.org/10.5281/zenodo.8411004. Source data are provided with this paper.
Code availability
MESuSiE (1.0) is available at https://github.com/borangao/MESuSiE, deposited at Zenodo55 https://doi.org/10.5281/zenodo.8411004. SuSiE (0.11.84) software is available at https://github.com/stephenslab/susieR. Paintor (3.0) software is available at https://github.com/gkichaev/PAINTOR_V3.0. METAL (metal-2011-03-25) is available at https://csg.sph.umich.edu/abecasis/metal/download/. Genotype data processing and quality control filtering of plink bed/fam/bim files were performed using PLINK (2.0) available at https://www.cog-genomics.org/plink/2.0/. Variants were annotated using Ensembl Variant Effect Predictor v85 with assembly GRCh37 is available at https://useast.ensembl.org/info/docs/tools/vep/index.html. Phylop score of variants are retrieved by R package GenomicScores (2.8.0). Base pair position of GTEX eQTL was transferred from GRCh38 to GRCh37 by R package liftOver (1.20.0). The analysis code to reproduce the results presented in the study is available at Zenodo55 https://doi.org/10.5281/zenodo.8411004.
References
Willer, C. J. et al. Newly identified loci that influence lipid concentrations and risk of coronary artery disease. Nat. Genet 40, 161–169 (2008).
Nielsen, J. B. et al. Genome-wide study of atrial fibrillation identifies seven risk loci and highlights biological pathways and regulatory elements involved in cardiac development. Am. J. Hum. Genet 102, 103–115 (2018).
Schaid, D. J. et al. From genome-wide associations to candidate causal variants by statistical fine-mapping. Nat. Rev. Genet. 19, 491–504 (2018).
Chen, W. et al. Fine mapping causal variants with an approximate Bayesian method using marginal test statistics. Genetics 200, 719–736 (2015).
Yang, J., Fritsche, L. G., Zhou, X. & Abecasis, G. A scalable Bayesian method for integrating functional information in genome-wide association studies. Am. J. Hum. Genet 101, 404–416 (2017).
Wen, X., Lee, Y., Luca, F. & Pique-Regi, R. Efficient integrative multi-SNP association analysis via deterministic approximation of posteriors. Am. J. Hum. Genet 98, 1114–1129 (2016).
Newcombe, P. J., Conti, D. V. & Richardson, S. JAM: a scalable Bayesian framework for joint analysis of marginal SNP effects. Genet. Epidemiol. 40, 188–201 (2016).
Benner, C. et al. FINEMAP: efficient variable selection using summary data from genome-wide association studies. Bioinformatics 32, 1493–1501 (2016).
Zhang, W., Najafabadi, H. S. & Li, Y. SparsePro: an efficient genome-wide fine-mapping method integrating summary statistics and functional annotations. Preprint at bioRxiv https://doi.org/10.1101/2021.10.04.463133 (2021).
Zou, Y., Carbonetto, P., Wang, G. & Stephens, M. Fine-mapping from summary data with the “Sum of Single Effects” model. PLoS Genet. 18, e1010299 (2022).
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. Ser. B Stat. Methodol. 82, 1273–1300 (2020).
Kichaev, G. et al. Integrating functional data to prioritize causal variants in statistical fine-mapping studies. PLoS Genet 10, e1004722 (2014).
Yang, Z. et al. CARMA is a new Bayesian model for fine-mapping in genome-wide association meta-analyses. Nat. Genet 55, 1057–1065 (2023).
Shi, H. et al. Localizing components of shared transethnic genetic architecture of complex traits from GWAS summary data. Am. J. Hum. Genet 106, 805–817 (2020).
Shi, H. et al. Population-specific causal disease effect sizes in functionally important regions impacted by selection. Nat. Commun. 12, 1–15 (2021).
Graham, S. E. et al. The power of genetic diversity in genome-wide association studies of lipids. Nature 600, 675–679 (2021).
Mahajan, A. et al. Genome-wide trans-ancestry meta-analysis provides insight into the genetic architecture of type 2 diabetes susceptibility. Nat. Genet 46, 234–244 (2014).
Wojcik, G. L. et al. Genetic analyses of diverse populations improves discovery for complex traits. Nature 570, 514–518 (2019).
Peterson, R. E. et al. Genome-wide association studies in ancestrally diverse populations: opportunities, methods, pitfalls, and recommendations. Cell 179, 589–603 (2019).
LaPierre, N. et al. Identifying causal variants by fine mapping across multiple studies. PLoS Genet 17, e1009733 (2021).
Mägi, R. et al. Trans-ethnic meta-regression of genome-wide association studies accounting for ancestry increases power for discovery and improves fine-mapping resolution. Hum. Mol. Genet 26, 3639–3650 (2017).
Kichaev, G. & Pasaniuc, B. Leveraging functional-annotation data in trans-ethnic fine-mapping studies. Am. J. Hum. Genet 97, 260–271 (2015).
Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203–209 (2018).
Sudlow, C. et al. UK Biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12, e10017779 (2015).
Benjamin, M. UK Biobank — Neale lab. http://www.nealelab.is/uk-biobank (2018).
Yuan, K. et al. Fine-mapping across diverse ancestries drives the discovery of putative causal variants underlying human complex traits and diseases. Preprint at medRxiv https://doi.org/10.1101/2023.01.07.23284293 (2023).
Zou, Y., Carbonetto, P., Xie, D., Wang, G. & Stephens, M. Fast and flexible joint fine-mapping of multiple traits via the Sum of Single Effects model. Preprint at bioRxiv https://doi.org/10.1101/2023.04.14.536893 (2023).
Weissbrod, O. et al. Functionally informed fine-mapping and polygenic localization of complex trait heritability. Nat. Genet 52, 1355–1363 (2020).
Kanai, M. et al. Insights from complex trait fine-mapping across diverse populations. Preprint at medRxiv https://doi.org/10.1101/2021.09.03.21262975 (2021).
Liang, Y. et al. A scalable unified framework of total and allele-specific counts for cis-QTL, fine-mapping, and prediction. Nat. Commun. 12, 1424 (2021).
Wang, Q. S. et al. Leveraging supervised learning for functionally informed fine-mapping of cis-eQTLs identifies an additional 20,913 putative causal eQTLs. Nat. Commun. 12, 3394 (2021).
Qiao, J., Shao, Z., Wu, Y., Zeng, P. & Wang, T. Detecting associated genes for complex traits shared across East Asian and European populations under the framework of composite null hypothesis testing. J. Transl. Med 20, 424 (2022).
Shang, L. et al. Genetic architecture of gene expression in European and African Americans: an eQTL mapping study in GENOA. Am. J. Hum. Genet 106, 496–512 (2020).
Nakamura, M. T. & Nara, T. Y. Structure, function, and dietary regulation of delta6, delta5, and delta9 desaturases. Annu Rev. Nutr. 24, 345–376 (2004).
Stoffel, W. et al. Obesity resistance and deregulation of lipogenesis in Δ6-fatty acid desaturase (FADS2) deficiency. EMBO Rep. 15, 110–120 (2014).
Nakaya, Y., Schaefer, E. J. & Brewer, H. B. Activation of human post heparin lipoprotein lipase by apolipoprotein H (β2-glycoprotein I). Top. Catal. 95, 1168–1172 (1980).
Choudhury, A. et al. Meta-analysis of sub-Saharan African studies provides insights into genetic architecture of lipid traits. Nat. Commun. 13, 2578 (2022).
Cavalcante, L. N. et al. African genetic ancestry is associated with lower frequency of PNPLA3 G allele in non-alcoholic fatty liver in an admixed population. Ann. Hepatol. 27, 100728 (2022).
Goffredo, M. et al. Role of TM6SF2 rs58542926 in the pathogenesis of nonalcoholic pediatric fatty liver disease: a multiethnic study. Hepatology 63, 117–125 (2016).
Fan, Y. et al. Hepatic transmembrane 6 superfamily member 2 regulates cholesterol metabolism in mice. Gastroenterology 150, 1208–1218 (2016).
O’Hare, E. A. et al. TM6SF2 rs58542926 impacts lipid processing in liver and small intestine. Hepatology 65, 1526–1542 (2017).
Zeng, P., Hao, X. & Zhou, X. Pleiotropic mapping and annotation selection in genome-wide association studies with penalized Gaussian mixture models. Bioinformatics 34, 2797–2807 (2018).
Zhou, G., Chen, T. & Zhao, H. SDPRX: a statistical method for cross-population prediction of complex traits. Am. J. Hum. Genet 110, 13–22 (2023).
Thompson, W. D. et al. Association of maternal circulating 25(OH)D and calcium with birth weight: a mendelian randomisation analysis. PLoS Med 16, e1002828 (2019).
Burns, A. C. et al. Genome-wide gene by environment study of time spent in daylight and chronotype identifies emerging genetic architecture underlying light sensitivity. Sleep 46, zsac287 (2023).
Gharahkhani, P. et al. Effect of increased body mass index on risk of diagnosis or death from cancer. Br. J. Cancer 120, 565–570 (2019).
Bovijn, J. et al. GWAS identifies risk locus for erectile dysfunction and implicates hypothalamic neurobiology and diabetes in etiology. Am. J. Hum. Genet 104, 157–163 (2019).
Yang, S. & Zhou, X. PGS-server: accuracy, robustness and transferability of polygenic score methods for biobank scale studies. Brief. Bioinform. 10, bbac039 (2022).
Horton, R. et al. Gene map of the extended human MHC. Nat. Genet. 5, 889–899 (2004).
Beck, S. et al. Complete sequence and gene map of a human major histocompatibility complex. Nature 401, 921–923 (1999).
Lonsdale, J. et al. The Genotype-Tissue Expression (GTEx) project. Nat. Genet. 45, 580–585 (2013).
Aguet, F. et al. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science 369, 1318–1330 (2020).
McLaren, W. et al. The Ensembl Variant Effect Predictor. Genome Biol. 17, 1–14 (2016).
Finucane, H. K. et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat. Genet 47, 1228–1235 (2015).
Gao, B. & Zhou, X. MESuSiE: enables scalable and powerful multi-ancestry fine-mapping of causal variants in genome-wide association studies. Zenodo https://doi.org/10.5281/zenodo.8411004 (2023).
Acknowledgements
This study was supported by the National Institutes of Health (NIH) grants R01HG009124 and R01GM144960 (to X.Z.). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. This study has been conducted using UK Biobank resource under Application Number 30686. UK Biobank was established by the Wellcome Trust medical charity, Medical Research Council, Department of Health, Scottish Government and the Northwest Regional Development Agency. It has also had funding from the Welsh Assembly Government, British Heart Foundation and Diabetes UK.
Author information
Authors and Affiliations
Contributions
X.Z. designed the methods. B.G. performed the experiments and analyzed and interpreted data. X.Z. and B.G. drafted and revised the manuscript. All authors critically reviewed the manuscript, suggested revisions as needed and approved the final version.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Genetics thanks Stephen Rich, Alicia Martin, Qiongshi Lu for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Comparison of the power of different methods in detecting the ancestry-specific causal signals in the baseline simulation settings.
Results are shown for the baseline setting with different numbers of causal SNPs (1, 3, or 5; columns) and different causal effect sizes (per causal SNP heritability = 10−4 or 2*10−4; rows). (a) Precision-recall curve with the PIP threshold spanning from 0.01 to 1. For each PIP threshold t, the White British-specific signal is declared as PIPWB > t for MESuSiE, PIPWB > t, PIPBB < t for SuSiE, and PIP > t for Paintor. the Black British-specific signal is declared as PIPBB > t for MESuSiE, PIPBB > t, PIPWB < t for SuSiE, and PIP > t for Paintor. These quantities are calculated as Precision = 1 – FDR, \({\rm{FDR}}\,=\frac{{\rm{FP}}}{{\rm{TP}}+{\rm{FP}}}\), \({\rm{Recall}}=\frac{{\rm{TP}}}{{\rm{TP}}+{\rm{FN}}}\), where FP, TP, FN, and TN denote the counts of false positives, true positives, false negatives, and true negatives respectively, at a specified PIP threshold. PIP threshold of 0.5 is labelled in circle. (b) Bar plot displays the power at controlled FDR of 0.01, 0.05, and 0.1.
Extended Data Fig. 2 Calibration of PIP detecting the ancestry-specific causal signals in the baseline simulation settings.
Results are shown for the baseline setting in n = 200 independent regions of interest (combining per causal SNP heritability = 10−4 or 2*10−4) with different number of causal signals (number of causal SNP = 1, 3, 5; cols). SNPs are ranked by PIPs and divided into 10 equally spaced bins based on PIP values. The observed proportion of the signal (y-axis) is compared to the mean PIP (x-axis) within each bin. The gray error bars in each panel represent 2 * standard errors. The diagonal red line in each panel shows the expected line when the PIP statistics are calibrated. Points below the diagonal line imply that the corresponding PIPs are larger than expected, suggesting more false discoveries than expected. Points above the diagonal line imply that the PIPs are conservative.
Extended Data Fig. 3 LocusZoom plot displays the fine-mapping results on TG from different methods in a genomic region on chromosome 19.
1st row: LocusZoom plots display the two-sided negative log10(P-value) from the marginal GWAS analysis (y-axis) in the UKBB (left) and the African ancestry GWAS (right) across base pair positions (x-axis). A genome-wide significance threshold of 5*10−8 was applied to determine marginal significance. Color of a SNP represents the r2 values between the SNP and the lead variant rs1801689 in the region. 2nd row: LocusZoom plots display the SuSiE PIP (y-axis) fitted in the UKBB (left) and the African ancestry GWAS (right) across base pair positions. 3rd row: LocusZoom plots for the PIPeither from MESuSiE (left) or Paintor (right) across base pair positions (x-axis). Signals, whether shared or ancestry-specific, are determined based on corresponding PIP value (PIPshared, PIPEUR, or PIPAFR) exceeding 0.5. 4th row: annotated genes in the genomic region. For the detected signals, we use upper triangle to represent European-specific signal, lower triangle to represent African ancestry-specific variant, diamond to represent shared signal, and square to represent Paintor signal. In contrast, a circle is used to indicate the absence of a signal (None). TG: triglycerides.
Extended Data Fig. 4 Comparison of the 95% credible sets from different methods for MCHC and SCZ.
(a) Boxplot displays the set size of the 95% credible sets for 174 analyzed genomic regions. (b) eQTL enrichment, which is calculated as the ratio of independent eSNP proportion within the 95% credible set to the eSNP proportion outside of the set. Blood and brain cortex eQTLs are used. MCHC: mean corpuscular hemoglobin concentration, SCZ: schizophrenia.
Extended Data Fig. 5 LocusZoom plot displays the fine-mapping results on SCZ from different methods in a genomic region on chromosome 15.
1st row: LocusZoom plots display the two-sided negative log10(P-value) from the marginal GWAS analysis (y-axis) in the PGC European (left) and the East Asian ancestry GWAS (right) across base pair positions (x-axis). A genome-wide significance threshold of 5*10−8 was applied to determine marginal significance. Color of a SNP represents the r2 values between the SNP and the lead variant rs4702 in the region. 2nd row: LocusZoom plots display the SuSiE PIP (y-axis) fitted in the European (left) and the East Asian ancestry GWAS (right) across base pair positions. 3rd row: LocusZoom plots for the PIPeither from MESuSiE (left) or Paintor (right) across base pair positions (x-axis). Signals, whether shared or ancestry-specific, are determined based on corresponding PIP value (PIPshared, PIPEUR, or PIPEAS) exceeding 0.5. 4th row: annotated genes in the genomic region. For the detected signals, we use upper triangle to represent European-specific signal, lower triangle to represent East Asian ancestry-specific variant, diamond to represent shared signal, and square to represent Paintor signal. In contrast, a circle is used to indicate the absence of a signal (None). SCZ: schizophrenia.
Supplementary information
Supplementary Information
Supplementary Figures 1–47, Supplementary Tables 1–4, Supplementary Note and References.
Source data
Source Data Fig. 2
Statistical source data.
Source Data Fig. 3
Statistical source data.
Source Data Fig. 4
Statistical source data.
Source Data Fig. 5
Statistical source data.
Source Data Fig. 6
Statistical source data.
Source Data Extended Data Fig./Table 1
Statistical source data.
Source Data Extended Data Fig./Table 2
Statistical source data.
Source Data Extended Data Fig./Table 3
Statistical source data.
Source Data Extended Data Fig./Table 4
Statistical source data.
Source Data Extended Data Fig./Table 5
Statistical source data.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Gao, B., Zhou, X. MESuSiE enables scalable and powerful multi-ancestry fine-mapping of causal variants in genome-wide association studies. Nat Genet 56, 170–179 (2024). https://doi.org/10.1038/s41588-023-01604-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41588-023-01604-7
