Perspective | Published:

Advantages and pitfalls in the application of mixed-model association methods

Nature Genetics volume 46, pages 100106 (2014) | Download Citation


Mixed linear models are emerging as a method of choice for conducting genetic association studies in humans and other organisms. The advantages of the mixed-linear-model association (MLMA) method include the prevention of false positive associations due to population or relatedness structure and an increase in power obtained through the application of a correction that is specific to this structure. An underappreciated point is that MLMA can also increase power in studies without sample structure by implicitly conditioning on associated loci other than the candidate locus. Numerous variations on the standard MLMA approach have recently been published, with a focus on reducing computational cost. These advances provide researchers applying MLMA methods with many options to choose from, but we caution that MLMA methods are still subject to potential pitfalls. Here we describe and quantify the advantages and pitfalls of MLMA methods as a function of study design and provide recommendations for the application of these methods in practical settings.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.


  1. 1.

    et al. A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat. Genet. 38, 203–208 (2006).

  2. 2.

    et al. An Arabidopsis example of association mapping in structured samples. PLoS Genet. 3, e4 (2007).

  3. 3.

    et al. Efficient control of population structure in model organism association mapping. Genetics 178, 1709–1723 (2008).

  4. 4.

    et al. Variance component model to account for sample structure in genome-wide association studies. Nat. Genet. 42, 348–354 (2010).

  5. 5.

    et al. Mixed linear model approach adapted for genome-wide association studies. Nat. Genet. 42, 355–360 (2010).

  6. 6.

    , , & New approaches to population stratification in genome-wide association studies. Nat. Rev. Genet. 11, 459–463 (2010).

  7. 7.

    et al. FaST linear mixed models for genome-wide association studies. Nat. Methods 8, 833–835 (2011).

  8. 8.

    et al. Improved linear mixed models for genome-wide association studies. Nat. Methods 9, 525–526 (2012).

  9. 9.

    et al. An efficient multi-locus mixed-model approach for genome-wide association studies in structured populations. Nat. Genet. 44, 825–830 (2012).

  10. 10.

    et al. A mixed-model approach for genome-wide association studies of correlated traits in structured populations. Nat. Genet. 44, 1066–1071 (2012).

  11. 11.

    & Genome-wide efficient mixed-model analysis for association studies. Nat. Genet. 44, 821–824 (2012).

  12. 12.

    , , , & Rapid variance components–based method for whole-genome association analysis. Nat. Genet. 44, 1166–1170 (2012).

  13. 13.

    et al. Common SNPs explain a large proportion of the heritability for human height. Nat. Genet. 42, 565–569 (2010).

  14. 14.

    & Heritability in the genome-wide association era. Hum. Genet. 131, 1655–1664 (2012).

  15. 15.

    Best linear unbiased estimation and prediction under a selection model. Biometrics 31, 423–447 (1975).

  16. 16.

    , & Predicting genetic predisposition in humans: the promise of whole-genome markers. Nat. Rev. Genet. 11, 880–886 (2010).

  17. 17.

    & Mixed models can correct for population structure for genomic regions under selection. Nat. Rev. Genet. 14, 300 (2013).

  18. 18.

    , , & Response to Sul and Eskin. Nat. Rev. Genet. 14, 300 (2013).

  19. 19.

    , & An analytical comparison of the principal component method and the mixed effects model for association studies in the presence of cryptic relatedness and population stratification. Hum. Hered. 76, 1–9 (2013).

  20. 20.

    & Differential confounding of rare and common variants in spatially structured populations. Nat. Genet. 44, 243–246 (2012).

  21. 21.

    & Family-based association tests for genomewide association scans. Am. J. Hum. Genet. 81, 913–926 (2007).

  22. 22.

    et al. GCTA: a tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82 (2011).

  23. 23.

    et al. Genetic risk and a primary role for cell-mediated immune mechanisms in multiple sclerosis. Nature 476, 214–219 (2011).

  24. 24.

    et al. Genomic inflation factors under polygenic inheritance. Eur. J. Hum. Genet. 19, 807–812 (2011).

  25. 25.

    et al. Genome partitioning of genetic variation for complex traits using common SNPs. Nat. Genet. 43, 519–525 (2011).

  26. 26.

    et al. The benefits of selecting phenotype-specific variants for applications of mixed models in genomics. Sci. Rep. 3, 1815 (2013).

  27. 27.

    et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat. Genet. 38, 904–909 (2006).

  28. 28.

    , & FaST-LMM-Select for addressing confounding from spatial structure and rare variants. Nat. Genet. 45, 470–471 (2013).

  29. 29.

    & The Covariate's Dilemma. PLoS Genet. 8, e1003096 (2012).

  30. 30.

    et al. Analysis of case-control association studies with known risk variants. Bioinformatics 28, 1729–1737 (2012).

  31. 31.

    Link functions in multi-locus genetic models: implications for testing, prediction, and interpretation. Genet. Epidemiol. 36, 409–418 (2012).

  32. 32.

    , & Including known covariates can reduce power to detect genetic effects in case-control studies. Nat. Genet. 44, 848–851 (2012).

  33. 33.

    et al. Informed conditioning on clinical covariates increases power in case-control association studies. PLoS Genet. 8, e1003032 (2012).

  34. 34.

    The inheritance of liability to diseases with variable age of onset, with particular reference to diabetes mellitus. Ann. Hum. Genet. 31, 1–20 (1967).

  35. 35.

    , , & Estimating missing heritability for disease from genome-wide association studies. Am. J. Hum. Genet. 88, 294–305 (2011).

  36. 36.

    et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature 491, 119–124 (2012).

  37. 37.

    et al. Estimation and partitioning of polygenic variation captured by common SNPs for Alzheimer's disease, multiple sclerosis and endometriosis. Hum. Mol. Genet. 22, 832–841 (2013).

  38. 38.

    & Genomic control for association studies. Biometrics 55, 997–1004 (1999).

  39. 39.

    et al. Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature 467, 832–838 (2010).

  40. 40.

    , & Prediction of total genetic value using genome-wide dense marker maps. Genetics 157, 1819–1829 (2001).

  41. 41.

    et al. Improving accuracy of genomic predictions within and between dairy cattle breeds with imputed high-density single nucleotide polymorphism panels. J. Dairy Sci. 95, 4114–4129 (2012).

  42. 42.

    , & Polygenic modeling with bayesian sparse linear mixed models. PLoS Genet. 9, e1003264 (2013).

Download references


We are grateful to N. Patterson, D. Heckerman, J. Listgarten, C. Lippert, E. Eskin, B. Vilhjalmsson, P. Loh, T. Hayeck, T. Frayling, A. McRae, L. Ronnegard, O. Weissbrod, G. Tucker and the GIANT Consortium for helpful discussions and to A. Gusev and S. Pollack for assistance with the multiple sclerosis and ulcerative colitis data sets. We are grateful to two anonymous referees for their helpful comments. This study makes use of data generated by the Wellcome Trust Case Control Consortium and data from the database of Genotypes and Phenotypes (dbGaP) under accessions phs000090.v2.p1 and phs000091.v2.p1 (see the Supplementary Note for the full set of acknowledgments for these data). This research was supported by US National Institutes of Health (NIH) grants R01 HG006399, P01 GM099568 and R01 GM075091, by the Australian Research Council (DP130102666) and by the Australian National Health and Medical Research Council (APP1011506 and APP1052684).

Author information

Author notes

    • Jian Yang
    •  & Noah A Zaitlen

    These authors contributed equally to this work.

    • Michael E Goddard
    • , Peter M Visscher
    •  & Alkes L Price

    These authors jointly directed this work.


  1. Queensland Brain Institute, University of Queensland, Brisbane, Queensland, Australia.

    • Jian Yang
    •  & Peter M Visscher
  2. University of Queensland Diamantina Institute, University of Queensland, Princess Alexandra Hospital, Brisbane, Queensland, Australia.

    • Jian Yang
    •  & Peter M Visscher
  3. Department of Medicine, Lung Biology Center, University of California, San Francisco, San Francisco, California, USA.

    • Noah A Zaitlen
  4. Faculty of Land and Food Resources, University of Melbourne, Parkville, Victoria, Australia.

    • Michael E Goddard
  5. Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, USA.

    • Alkes L Price
  6. Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts, USA.

    • Alkes L Price
  7. Program in Medical and Population Genetics, Broad Institute of MIT and Harvard, Cambridge, Massachusetts, USA.

    • Alkes L Price


  1. Search for Jian Yang in:

  2. Search for Noah A Zaitlen in:

  3. Search for Michael E Goddard in:

  4. Search for Peter M Visscher in:

  5. Search for Alkes L Price in:


All authors conceived the project and designed the analyses. J.Y., N.A.Z. and A.L.P. performed the analyses. J.Y., M.E.G. and P.M.V. provided the theoretical derivations. J.Y. wrote the GCTA software. J.Y., N.A.Z. and A.L.P. wrote the manuscript with edits from all authors.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Peter M Visscher or Alkes L Price.

Supplementary information

PDF files

  1. 1.

    Supplementary Text and Figures

    Supplementary Figure 1, Supplementary Tables 1–11 and Supplementary Note.

About this article

Publication history





Further reading