Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Correspondence
  • Published:

Abundant associations with gene expression complicate GWAS follow-up

Nature Genetics volume 51, pages 768–769 (2019)Cite this article

Subjects

To the Editor — Genome-wide association studies (GWAS) are rapidly expanding the catalog of trait- and disease-associated variants. With increasing cohort size and phenotyping, GWAS have identified more than 70,000 associated variants1. Because as many as 90% of GWAS variants fall within non-coding regions, most of them have unknown functional importance2. To aid in interpreting these variants, expression quantitative trait locus (eQTL) studies provide data on whether a variant of interest is also associated with gene expression levels.

To improve GWAS follow-up, we developed an online platform called LocusCompare to facilitate the visualization of colocalization events. We integrated into the web server more than 200 peer-reviewed GWAS studies across more than 800 unique traits and 642 disease-associated phenotypes from the UK Biobank rapid GWAS (downloaded from http://www.nealelab.is/uk-biobank/). In addition, LocusCompare integrates eQTLs from 48 tissues in the GTEx study (version 7)3; eQTLs and splicing QTLs from coronary artery smooth muscle cells9 and retinal pigment epithelial cells10; and methylation QTLs from brain tissues11,12 and whole blood13. Using preloaded association datasets, LocusCompare enables easy comparison between pairs of association signals. Although colocalization analyses between GWAS and eQTL are the most common, LocusCompare also enables comparison between two GWAS or two eQTL datasets to detect pleiotropy (Supplementary Fig. 3). Currently, stacked Manhattan plots are the most frequently used strategy to visualize colocalization of association signals14. However, such a visualization strategy could mistake nearby variants in low LD as shared lead variants in a colocalization event (Supplementary Fig. 4a). To mitigate such confounding, we introduce a modified scatter plot (the LocusCompare plot) to visualize colocalization events (Fig. 1b,c). Each dot represents a variant and is colored according to its LD to the selected variant. A bona fide colocalization signal should form a single spike toward the top right corner, as illustrated by the well-known colocalization between SORT1 eQTL in the liver and coronary artery disease GWAS (Supplementary Fig. 5). To enable exploration of GWAS–eQTL colocalization, we performed colocalization analyses6,15 across all loci with GWAS P value <5 × 10–8 and eQTL P value <1 × 10–6 for studies hosted on the web server (Supplementary Methods). Users can visualize all tested genes with a Manhattan plot for any given GWAS and eQTL colocalization, and can click on promising genes for further investigation. In addition, LocusCompare is highly extensible in that it allows users to upload custom association datasets and visualize within the LocusCompare web framework. To accommodate advanced usage, we provide an R package, LocusCompareR, for visualization of colocalization events in local environments and a bash script to download all curated GWAS studies.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy this article

  • Purchase on SpringerLink
  • Instant access to the full article PDF.

USD 39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Distinguishing candidates from false-positive genes by using LocusCompare.

Data availability

A bash script to download all GWAS datasets is available at https://github.com/mikegloudemans/gwas-download. GTEx eQTL data can be accessed via https://gtexportal.org/home/datasets. Data for coronary artery smooth muscle cells are available at https://stanford.app.box.com/s/e6e8hyft5u7wix1nzg5mjfqa084c4tin. Data for retinal pigment epithelium cells are available at https://stanford.box.com/s/asrxy0o66xxe1j7mfj56p3z3d405gijj. Methylation QTL data for brain and blood are available at https://cnsgenomics.com/software/smr/#DataResource.

Code availability

LocusCompare is hosted at http://locuscompare.com and is also available as open source software at https://github.com/boxiangliu/locuscompare. LocusCompareR is open source and is available at https://github.com/boxiangliu/locuscomparer. Our pipeline for running colocalization tests is available at https://bitbucket.org/mgloud/production_coloc_pipeline/. The LocusCompareR package was built with R version 3.2.3.

References

  1. MacArthur, J. et al. Nucleic Acids Res. 45, D896–D901 (2017).

    Article  CAS  Google Scholar 

  2. Farh, K. K.-H. et al. Nature 518, 337–343 (2015).

    Article  CAS  Google Scholar 

  3. Battle, A., Brown, C. D., Engelhardt, B. E. & Montgomery, S. B. Nature 550, 204–213 (2017).

    Article  Google Scholar 

  4. Zhao, W. et al. Nat. Genet. 49, 1450–1457 (2017).

    Article  CAS  Google Scholar 

  5. Nica, A. C. et al. PLoS Genet. 6, e1000895 (2010).

    Article  Google Scholar 

  6. Hormozdiari, F. et al. Am. J. Hum. Genet. 99, 1245–1260 (2016).

    Article  CAS  Google Scholar 

  7. Zhu, Z. et al. Nat. Genet. 48, 481–487 (2016).

    Article  CAS  Google Scholar 

  8. Giambartolomei, C. et al. PLoS Genet. 10, e1004383 (2014).

    Article  Google Scholar 

  9. Liu, B. et al. Am. J. Hum. Genet. 103, 377–388 (2018).

    Article  CAS  Google Scholar 

  10. Liu, B. et al. Preprint at bioRxiv https://doi.org/10.1101/446799 (2018).

  11. Qi, T. et al. Nat. Commun. 9, 2282 (2018).

    Article  Google Scholar 

  12. Hannon, E. et al. Nat. Neurosci. 19, 48–54 (2016).

    Article  CAS  Google Scholar 

  13. McRae, A. F. et al. Sci. Rep. 8, 17605 (2018).

    Article  Google Scholar 

  14. Pruim, R. J. et al. Bioinformatics 26, 2336–2337 (2010).

    Article  CAS  Google Scholar 

  15. Benner, C. et al. Bioinformatics 32, 1493–1501 (2016).

    Article  CAS  Google Scholar 

Download references

Acknowledgements

B.L. is supported by the Stanford Center for Evolution and Human Genomics fellowship and National Key R&D Program of China, 2016YFD0400800 and Baidu Research. M.J.G. is funded by NLM training grant T15 LM 007033 and a Stanford Graduate Fellowship. E.I. is supported by R01DK106236. S.B.M. is supported by R33HL120757 (NHLBI), U01HG009431 (NHGRI; ENCODE4), R01MH101814 (NIH Common Fund; GTEx Program), R01HG008150 (NHGRI; Non-Coding Variants Program), R01HL142015 (NHLBI; TOPMED), U01HG009080 (NHGRI; GSPAC) and the Edward Mallinckrodt Jr. Foundation. We acknowledge A. Shcherbina for support in this project and N. Cyr for support with graphical illustration.

Author information

Author notes

  1. These authors contributed equally: Boxiang Liu, Michael J. Gloudemans.

Authors and Affiliations

  1. Department of Biology, Stanford University, Stanford, CA, USA

    Boxiang Liu

  2. Department of Pathology, Stanford University School of Medicine, Stanford, CA, USA

    Boxiang Liu, Michael J. Gloudemans, Abhiram S. Rao & Stephen B. Montgomery

  3. Biomedical Informatics Training Program, Stanford University School of Medicine, Stanford, CA, USA

    Michael J. Gloudemans

  4. Department of Bioengineering, Stanford University, Stanford, CA, USA

    Abhiram S. Rao

  5. Department of Medicine, Division of Cardiovascular Medicine, Stanford University School of Medicine, Stanford, CA, USA

    Erik Ingelsson

  6. Stanford Cardiovascular Institute, Stanford University, Stanford, CA, USA

    Erik Ingelsson

  7. Stanford Diabetes Research Center, Stanford University, Stanford, CA, USA

    Erik Ingelsson

  8. Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA

    Stephen B. Montgomery

Authors
  1. Boxiang Liu
    View author publications

    Search author on:PubMed Google Scholar

  2. Michael J. Gloudemans
    View author publications

    Search author on:PubMed Google Scholar

  3. Abhiram S. Rao
    View author publications

    Search author on:PubMed Google Scholar

  4. Erik Ingelsson
    View author publications

    Search author on:PubMed Google Scholar

  5. Stephen B. Montgomery
    View author publications

    Search author on:PubMed Google Scholar

Contributions

B.L. and S.B.M. designed the study. B.L., M.J.G. and A.S.R. performed analyses. E.I. provided study support and feedback. B.L., M.J.G. and S.B.M. wrote the paper. B.L. and M.J.G. contributed equally to this work.

Corresponding authors

Correspondence to Boxiang Liu or Stephen B. Montgomery.

Ethics declarations

Competing interests

S.B.M. is on the SAB of Prime Genomics.

Supplementary information

Supplementary Information

Supplementary Methods and Supplementary Figs. 1–6

Reporting Summary

Supplementary Tables 1 and 2

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Liu, B., Gloudemans, M.J., Rao, A.S. et al. Abundant associations with gene expression complicate GWAS follow-up. Nat Genet 51, 768–769 (2019). https://doi.org/10.1038/s41588-019-0404-0

Download citation

  • Published:

  • Version of record:

  • Issue date:

  • DOI: https://doi.org/10.1038/s41588-019-0404-0

This article is cited by

Search

Advanced search

Quick links

Nature Briefing: Translational Research

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

Get what matters in translational research, free to your inbox weekly. Sign up for Nature Briefing: Translational Research