Improved linear mixed models for genome-wide association studies (original) (raw)

Nature Methods volume 9, pages 525–526 (2012)Cite this article

Subjects

To the Editor:

The use of linear mixed models (LMMs) in genome-wide association studies (GWAS) is now widely accepted1 because LMMs have been shown to be capable of correcting for several forms of confounding due to genetic relatedness, such as population structure and familial relatedness1, and because recent advances have made them computationally efficient1,2. LMMs tackle confounding by using a matrix of pairwise genetic similarities to model the relatedness among subjects. The consensus until now has been that all available single-nucleotide polymorphisms (SNPs) should be used to determine these similarities1. Here, however, we show theoretically and experimentally that carefully selecting a small number of SNPs systematically increases power (that is, it jointly reduces false positives and false negatives), improves calibration (lessens inflation or deflation of the test statistic) and reduces computational cost.

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

Relevant articles

Open Access articles citing this article.

Access options

Subscribe to this journal

Receive 12 print issues and online access

$259.00 per year

only $21.58 per issue

Buy this article

USD 39.95

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

Additional access options:

References

  1. Kang, H.M. et al. Nat. Genet. 42, 348–354 (2010).
    Article CAS Google Scholar
  2. Lippert, C. et al. Nat. Methods 8, 833–835 (2011).
    Article CAS Google Scholar
  3. Hayes, B.J., Visscher, P.M. & Goddard, M.E. Genet. Res. (Camb.) 91, 47–60 (2009).
    Article CAS Google Scholar
  4. Hoggart, C.J., Whittaker, J.C., Iorio, M.D. & Balding, D.J. PLoS Genet. 4, e1000130 (2008).
    Article Google Scholar
  5. Setakis, E., Stirnadel, H. & Balding, D.J. Genome Res. 16, 290–296 (2006).
    Article CAS Google Scholar
  6. Wellcome Trust Case Control Consortium. Nature 447, 661–678 (2007).

Download references

Acknowledgements

We thank J. Carlson for help with tools to manage and analyze the data and P. Palamara for cataloging the positions and genetic distances of SNPs in the data for Crohn's disease. This study makes use of data generated by the Wellcome Trust Case Control Consortium. A full list of the investigators who contributed to the generation of the data is available from http://www.wtccc.org.uk/. Funding for the project was provided by the Wellcome Trust under award 076113 and 085475. E.E. is supported by US National Science Foundation grants 0916676 and 1065276 and by US National Institutes of Health grants K25-HL080079, U01-DA024417, P01-HL30568 and PO1-HL28481.

Author information

Author notes

  1. Jennifer Listgarten, Christoph Lippert and David Heckerman: These authors contributed equally to this work.

Authors and Affiliations

  1. Microsoft Research, Los Angeles, California, USA
    Jennifer Listgarten, Christoph Lippert & David Heckerman
  2. Max Planck Institutes Tübingen, Tübingen, Germany
    Christoph Lippert
  3. Microsoft Research, Redmond, Washington, USA
    Carl M Kadie & Robert I Davidson
  4. University of California Los Angeles, Los Angeles, California, USA
    Eleazar Eskin

Authors

  1. Jennifer Listgarten
  2. Christoph Lippert
  3. Carl M Kadie
  4. Robert I Davidson
  5. Eleazar Eskin
  6. David Heckerman

Corresponding authors

Correspondence toJennifer Listgarten, Christoph Lippert or David Heckerman.

Ethics declarations

Competing interests

J.L., C.L., C.M.K., R.I.D. and D.H. performed research related to this manuscript while employed by Microsoft.

Supplementary information

Rights and permissions

About this article

Cite this article

Listgarten, J., Lippert, C., Kadie, C. et al. Improved linear mixed models for genome-wide association studies.Nat Methods 9, 525–526 (2012). https://doi.org/10.1038/nmeth.2037

Download citation

This article is cited by