Skip to main content

Thank you for visiting 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.

Pitfalls of predicting complex traits from SNPs


The success of genome-wide association studies (GWASs) has led to increasing interest in making predictions of complex trait phenotypes, including disease, from genotype data. Rigorous assessment of the value of predictors is crucial before implementation. Here we discuss some of the limitations and pitfalls of prediction analysis and show how naive implementations can lead to severe bias and misinterpretation of results.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Flowchart of SNP-based prediction analysis.
Figure 2: The overlap pitfall: non-independence of discovery and validation samples.


  1. 1

    de los Campos, G., Gianola, D. & Allison, D. B. Predicting genetic predisposition in humans: the promise of whole-genome markers. Nature Rev. Genet. 11, 880–886 (2010).

    CAS  PubMed  Google Scholar 

  2. 2

    Gonzalez-Camacho, J. M. et al. Genome-enabled prediction of genetic values using radial basis function neural networks. Theor. Appl. Genet. 125, 759–771 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  3. 3

    Crossa, J. et al. Prediction of genetic values of quantitative traits in plant breeding using pedigree and molecular markers. Genetics 186, 713–724 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  4. 4

    Wei, Z. et al. From disease association to risk assessment: an optimistic view from genome-wide association studies on type 1 diabetes. PLoS Genet. 5, e1000678 (2009).

    PubMed  PubMed Central  Google Scholar 

  5. 5

    de los Campos, G., Hickey, J. M., Pong-Wong, R., Daetwyler, H. D. & Calus, M. P. L. Whole genome regression and prediction methods applied to plant and animal breeding. Genetics 193, 1255–1268 (2012).

    Google Scholar 

  6. 6

    Heffner, E. L., Sorrells, M. E. & Jannink, J. L. Genomic selection for crop improvement. Crop Sci. 49, 1–12 (2009).

    CAS  Google Scholar 

  7. 7

    Riedelsheimer, C. et al. Genomic and metabolic prediction of complex heterotic traits in hybrid maize. Nature Genet. 44, 217–220 (2012).

    CAS  Google Scholar 

  8. 8

    Becker, F. et al. Genetic testing and common disorders in a public health framework: how to assess relevance and possibilities. Eur. J. Hum. Genet. 19, S6–S44 (2011).

    PubMed  PubMed Central  Google Scholar 

  9. 9

    Visscher, P. M., Hill, W. G. & Wray, N. R. Heritability in the genomics era—concepts and misconceptions. Nature Rev. Genet. 9, 255–266 (2008).

    CAS  Google Scholar 

  10. 10

    Janssens, A. C. et al. Predictive testing for complex diseases using multiple genes: fact or fiction? Genet. Med. 8, 395–400 (2006).

    Google Scholar 

  11. 11

    Wray, N. R., Yang, J., Goddard, M. E. & Visscher, P. M. The genetic interpretation of area under the ROC curve in genomic profiling. PLoS Genet. 6, e1000864 (2010).

    PubMed  PubMed Central  Google Scholar 

  12. 12

    Burga, A., Casanueva, M. O. & Lehner, B. Predicting mutation outcome from early stochastic variation in genetic interaction partners. Nature 480, 250–253 (2011).

    CAS  Google Scholar 

  13. 13

    Seddon, J. M. et al. Prediction model for prevalence and incidence of advanced age-related macular degeneration based on genetic, demographic, and environmental variables. Invest. Ophthalmol. Vis. Sci. 50, 2044–2053 (2009).

    Google Scholar 

  14. 14

    Polychronakos, C. & Li, Q. Understanding type 1 diabetes through genetics: advances and prospects. Nature Rev. Genet. 12, 781–792 (2011).

    CAS  Google Scholar 

  15. 15

    So, H. C., Kwan, J. S., Cherny, S. S. & Sham, P. C. Risk prediction of complex diseases from family history and known susceptibility loci, with applications for cancer screening. Am. J. Hum. Genet. 88, 548–565 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  16. 16

    Pharoah, P. D., Antoniou, A. C., Easton, D. F. & Ponder, B. A. Polygenes, risk prediction, and targeted prevention of breast cancer. N. Engl. J. Med. 358, 2796–2803 (2008).

    CAS  Google Scholar 

  17. 17

    Chatterjee, N. et al. Projecting the performance of risk prediction based on polygenic analyses of genome-wide association studies. Nature Genet. 45, 400–405 (2013).

    CAS  Google Scholar 

  18. 18

    Tenesa, A. & Haley, C. S. The heritability of human disease: estimation, uses and abuses. Nature Rev. Genet. 14, 139–149 (2013).

    CAS  Google Scholar 

  19. 19

    Ayodo, G. et al. Combining evidence of natural selection with association analysis increases power to detect malaria-resistance variants. Am. J. Hum. Genet. 81, 234–242 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  20. 20

    Raj, T. et al. Alzheimer disease susceptibility loci: evidence for a protein network under natural selection. Am. J. Hum. Genet. 90, 720–726 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  21. 21

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

    CAS  PubMed  PubMed Central  Google Scholar 

  22. 22

    Barreiro, L. B., Laval, G., Quach, H., Patin, E. & Quintana-Murci, L. Natural selection has driven population differentiation in modern humans. Nature Genet. 40, 340–345 (2008).

    CAS  Google Scholar 

  23. 23

    Crow, J. F. Maintaining evolvability. J. Genet. 87, 349–353 (2008).

    Google Scholar 

  24. 24

    Vissers, L. E. et al. A de novo paradigm for mental retardation. Nature Genet. 42, 1109–1112 (2010).

    CAS  Google Scholar 

  25. 25

    de Brouwer, A. P. et al. Mutation frequencies of X-linked mental retardation genes in families from the EuroMRX consortium. Hum. Mutat. 28, 207–208 (2007).

    Google Scholar 

  26. 26

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

    CAS  PubMed  Google Scholar 

  27. 27

    Visscher, P. M. et al. A commentary on 'Common SNPs explain a large proportion of the heritability for human height' by Yang et al. (2010). Twin. Res. Hum. Genet. 13, 517–524 (2010).

    Google Scholar 

  28. 28

    Visscher, P. M., Brown, M. A., McCarthy, M. I. & Yang, J. Five years of GWAS discovery. Am. J. Hum. Genet. 90, 7–24 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  29. 29

    Purcell, S. M. et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460, 748–752 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. 30

    Lee, S. H. et al. Estimating the proportion of variation in susceptibility to schizophrenia captured by common SNPs. Nature Genet. 44, 247–250 (2012).

    CAS  Google Scholar 

  31. 31

    Haile-Mariam, M., Nieuwhof, G. J., Beard, K. T., Konstatinov, K. V. & Hayes, B. J. Comparison of heritabilities of dairy traits in Australian Holstein-Friesian cattle from genomic and pedigree data and implications for genomic evaluations. J. Anim. Breed. Genet. 130, 20–31 (2013).

    CAS  Google Scholar 

  32. 32

    Jensen, J., Su, G. & Madsen, P. Partitioning additive genetic variance into genomic and remaining polygenic components for complex traits in dairy cattle. BMC Genet. 13, 44 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  33. 33

    Kemper, K. E., Daetwyler, H. D., Visscher, P. M. & Goddard, M. E. Comparing linkage and association analyses in sheep points to a better way of doing GWAS. Genet. Res. 94, 191–203 (2012).

    CAS  Google Scholar 

  34. 34

    Wu, M. C. et al. Rare-variant association testing for sequencing data with the sequence kernel association test. Am. J. Hum. Genet. 89, 82–93 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  35. 35

    Bacanu, S. A., Nelson, M. R. & Whittaker, J. C. Comparison of statistical tests for association between rare variants and binary traits. PLoS ONE 7, e42530 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  36. 36

    Lindor, N. M. et al. A review of a multifactorial probability-based model for classification of BRCA1 and BRCA2 variants of uncertain significance (VUS). Hum. Mutat. 33, 8–21 (2012).

    CAS  Google Scholar 

  37. 37

    Stahl, E. A. et al. Bayesian inference analyses of the polygenic architecture of rheumatoid arthritis. Nature Genet. 44, 483–489 (2012).

    CAS  PubMed  Google Scholar 

  38. 38

    Goddard, M. E. Genomic selection: prediction of accuracy and maximisation of long term response. Genetica 136, 245–257 (2009).

    Google Scholar 

  39. 39

    Hayes, B. J., Bowman, P. J., Chamberlain, A. J. & Goddard, M. E. Invited review: Genomic selection in dairy cattle: progress and challenges. J. Dairy Sci. 92, 433–443 (2009).

    CAS  PubMed  Google Scholar 

  40. 40

    Daetwyler, H. D., Villanueva, B. & Woolliams, J. A. Accuracy of predicting the genetic risk of disease using a genome-wide approach. PLoS ONE 3, e3395 (2008).

    PubMed  PubMed Central  Google Scholar 

  41. 41

    de los Campos, G. et al. Predicting quantitative traits with regression models for dense molecular markers and pedigree. Genetics 182, 375–385 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. 42

    Goddard, M. E., Wray, N. R., Verbyla, K. L. & Visscher, P. M. Estimating effects and making predictions from genome-wide marker data. Statist. Sci. 24, 517–529 (2009).

    Google Scholar 

  43. 43

    Stephens, M. & Balding, D. J. Bayesian statistical methods for genetic association studies. Nature Rev. Genet. 10, 681–690 (2009).

    CAS  Google Scholar 

  44. 44

    Guan, Y. T. & Stephens, M. Bayesian variable selection regression for genome-wide association studies and other large-scale problems. Ann. Appl. Statist. 5, 1780–1815 (2011).

    Google Scholar 

  45. 45

    Zhou, X., Carbonetto, P. & Stephens, M. Polygenic modeling with Bayesian sparse linear mixed models. PLoS Genet. 9, e1003264 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  46. 46

    Erbe, M. 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).

    CAS  Google Scholar 

  47. 47

    Yang, J. et al. Conditional and joint multiple-SNP analysis of GWAS summary statistics identifies additional variants influencing complex traits. Nature Genet. 44, 369–375 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  48. 48

    Meigs, J. B. et al. Genotype score in addition to common risk factors for prediction of type 2 diabetes. N. Engl. J. Med. 359, 2208–2219 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  49. 49

    Kraft, P. & Hunter, D. J. Genetic risk prediction—are we there yet? N. Engl. J. Med. 360, 1701–1703 (2009).

    CAS  Google Scholar 

  50. 50

    Paynter, N. P. et al. Association between a literature-based genetic risk score and cardiovascular events in women. JAMA 303, 631–637 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  51. 51

    Wacholder, S. et al. Performance of common genetic variants in breast-cancer risk models. N. Engl. J. Med. 362, 986–993 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  52. 52

    Meuwissen, T. H., Hayes, B. J. & Goddard, M. E. Prediction of total genetic value using genome-wide dense marker maps. Genetics 157, 1819–1829 (2001).

    CAS  PubMed  PubMed Central  Google Scholar 

  53. 53

    Ober, U. et al. Using whole-genome sequence data to predict quantitative trait phenotypes in Drosophila melanogaster. PLoS Genet. 8, e1002685 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  54. 54

    Abraham, G., Kowalczyk, A., Zobel, J. & Inouye, M. SparSNP: fast and memory-efficient analysis of all SNPs for phenotype prediction. BMC Bioinformatics 13, 88 (2012).

    PubMed  PubMed Central  Google Scholar 

  55. 55

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

    PubMed  PubMed Central  Google Scholar 

  56. 56

    Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006).

    PubMed  PubMed Central  Google Scholar 

  57. 57

    Derringer, J. et al. Predicting sensation seeking from dopamine genes. A candidate-system approach. Psychol. Sci. 21, 1282–1290 (2010).

    PubMed  PubMed Central  Google Scholar 

  58. 58

    Mackay, T. F. et al. The Drosophila melanogaster Genetic Reference Panel. Nature 482, 173–178 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  59. 59

    Powell, J. E. & Zietsch, B. P. Predicting sensation seeking from dopamine genes: use and misuse of genetic prediction. Psychol. Sci. 22, 413–415 (2011).

    Google Scholar 

  60. 60

    Skafidas, E. et al. Predicting the diagnosis of autism spectrum disorder using gene pathway analysis. Mol. Psychiatry 11 Sep 2012 (10.1038/mp.2012.126).

  61. 61

    Ambroise, C. & McLachlan, G. J. Selection bias in gene extraction on the basis of microarray gene-expression data. Proc. Natl Acad. Sci. USA 99, 6562–6566 (2002).

    CAS  Google Scholar 

  62. 62

    Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81, 559–575 (2007).

    CAS  PubMed  PubMed Central  Google Scholar 

  63. 63

    Yang, J., Lee, S. H., Goddard, M. E. & Visscher, P. M. GCTA: a tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  64. 64

    Makowsky, R. et al. Beyond missing heritability: prediction of complex traits. PLoS Genet. 7, e1002051 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  65. 65

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

    CAS  PubMed  PubMed Central  Google Scholar 

  66. 66

    Daetwyler, H. D., Calus, M. P. L., Pong-Wong, R., de los Campos, G. & Hickey, J. M. Genomic prediction in animals and plants: simulation of data, validation, reporting and benchmarking. Genetics 193, 347–365 (2012).

    Google Scholar 

  67. 67

    Price, A. L. et al. Discerning the ancestry of European Americans in genetic association studies. PLoS Genet. 4, e236 (2008).

    PubMed  PubMed Central  Google Scholar 

  68. 68

    Belgard, T. G., Jankovic, I., Lowe, J. K. & Geschwind, D. H. Population structure confounds autism genetic classifier. Mol. Psychiatry 2 Apr 2013 (10.1038/mp.2013.34).

  69. 69

    Lee, S. H., Wray, N. R., Goddard, M. E. & Visscher, P. M. Estimating missing heritability for disease from genome-wide association studies. Am. J. Hum. Genet. 88, 294–305 (2011).

    PubMed  PubMed Central  Google Scholar 

  70. 70

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

    CAS  PubMed  Google Scholar 

  71. 71

    Thornton, T. et al. Estimating kinship in admixed populations. Am. J. Hum. Genet. 91, 122–138 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  72. 72

    Lubke, G. H. et al. Estimating the genetic variance of major depressive disorder due to all single nucleotide polymorphisms. Biol. Psychiatry 72, 707–709 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  73. 73

    Machiela, M. J. et al. Evaluation of polygenic risk scores for predicting breast and prostate cancer risk. Genet. Epidemiol. 35, 506–514 (2011).

    PubMed  PubMed Central  Google Scholar 

  74. 74

    Evans, D. M., Visscher, P. M. & Wray, N. R. Harnessing the information contained within genome-wide association studies to improve individual prediction of complex disease risk. Hum. Mol. Genet. 18, 3525–3531 (2009).

    CAS  Google Scholar 

  75. 75

    Peterson, R. E. et al. Genetic risk sum score comprised of common polygenic variation is associated with body mass index. Hum. Genet. 129, 221–230 (2011).

    Google Scholar 

  76. 76

    Lee, S. H., Yang, J., Goddard, M. E., Visscher, P. M. & Wray, N. R. Estimation of pleiotropy between complex diseases using single-nucleotide polymorphism-derived genomic relationships and restricted maximum likelihood. Bioinformatics 28, 2540–2542 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  77. 77

    Campbell, C. D. et al. Demonstrating stratification in a European American population. Nature Genet. 37, 868–872 (2005).

    CAS  Google Scholar 

  78. 78

    Turchin, M. C. et al. Evidence of widespread selection on standing variation in Europe at height-associated SNPs. Nature Genet. 44, 1015–1019 (2012).

    CAS  PubMed  PubMed Central  Google Scholar 

  79. 79

    Psaty, B. M. et al. Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium: design of prospective meta-analyses of genome-wide association studies from 5 cohorts. Circ. Cardiovasc. Genet. 2, 73–80 (2009).

    PubMed  PubMed Central  Google Scholar 

  80. 80

    Qi, L. et al. Genetic variants at 2q24 are associated with susceptibility to type 2 diabetes. Hum. Mol. Genet. 19, 2706–2715 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  81. 81

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

    CAS  PubMed  Google Scholar 

Download references


The authors acknowledge funding from the Australian National Health and Medical Research Council (1047956, 1011506, 613601, 613602, 1048853, 1052684, 1050218), Australian Research Council (FT0991360, DP130102666) and the US National Institutes of Health (NIH; R01 HG006399, R01 GM 075091, P01 GM 099568, R01 MH100141). The authors thank J. Witte for helpful comments. Funding support for the genome-wide association study (GWAS) of gene and environment initiatives in type 2 diabetes is provided through the NIH Genes, Environment and Health Initiative (GEI; U01HG004399). The human subjects participating in the GWAS are derived from the Nurses' Health Study (NHS) and Health Professionals' Follow-up Study (HPFS), and these studies are supported by NIH grants CA87969, CA55075 and DK58845. Assistance with phenotype harmonization and genotype cleaning, as well as with general study coordination, was provided by the gene–environment association studies, GENEVA Coordinating Center (U01 HG004446). Assistance with data cleaning was provided by the US National Center for Biotechnology Information (NCBI). Funding support for genotyping, which was carried out at the Broad Institute of MIT and Harvard, was provided by the NIH GEI (U01HG004424). The data sets used for the analyses described in this manuscript were obtained from dbGap accession number phs000091. The Atherosclerosis Risk in Communities Study (ARIC) was carried out as a collaborative study supported by US National Heart, Lung, and Blood Institute (NHLBI) contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, HHSN268201100012C), R01HL087641, R01HL59367 and R01HL086694; National Human Genome Research Institute contract U01HG004402; and NIH contract HHSN268200625226C. The authors thank the staff and participants of the ARIC study for their important contributions. Infrastructure was partly supported by grant number UL1RR025005, a component of the NIH and NIH Roadmap for Medical Research. The Framingham Heart Study (FHS) is conducted and supported by the NHLBI in collaboration with Boston University (Contract No. N01-HC-25195). This article was not prepared in collaboration with investigators of the FHS and does not necessarily reflect the opinions or views of the FHS, Boston University or the NHLBI. Funding for SHARe Affymetrix genotyping was provided by NHLBI Contract N02-HL-64278. SHARe Illumina genotyping was provided under an agreement between Illumina and Boston University. The authors are grateful to S. Pollack, C. Palmer and J. Hirschhorn for assistance with FHS data.

Author information



Corresponding author

Correspondence to Peter M. Visscher.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Related links

PowerPoint slides

Supplementary information

Supplementary Information S1 (Box)

Derivation of expected R2 when discovery data are used both for SNP selection and prediction (Box 2). (PDF 271 kb)

Supplementary Information S2 (table)

Analysis of height from Framingham Heart Study (Box 3) (PDF 185 kb)

Supplementary Information S3 (box)

Analysis of dairy cattle data (Fig 2a) (PDF 271 kb)


Ancestry principal components

Principal components derived from the genome relationship matrix that account for the genetic substructure of the data. In case–control studies, these principal components can reflect genotyping artefacts, such as plate, batch and genotyping centre, that could be confounded with case–control status.

Conventionally unrelated

Individuals that are not closely related: for example, more distantly related than third cousins.


Cross-validation involves testing the validity of a prediction in the absence of an independent external validation sample. This is done by dividing the sample into k independent subsets (balanced with respect to case–control status in disease data). Each of the k subsets is used in turn as a validation sample for a predictor derived from the remaining k– 1 subsets.

Cryptic relatedness

When a sample is thought to comprise unrelated individuals on the basis of recorded pedigree relationships but in fact includes close relatives: for example, second cousin or closer.

Effective population size

The number of individuals in an idealized population with random mating and no selection that would lead to the same rate of inbreeding as observed in the real population.

Estimated breeding values

Estimates of the additive genetic value for a particular trait that an individual will pass on to descendants.


The proportion of phenotypic variance attributable to additive genetic variation.

Independent sample

In the context of risk prediction, this is a sample from the same population but excluding individuals that are closely related. It is necessary for the individuals in different samples from the same population to share common ancestors, and indeed this distant sharing underpins the efficacy of a risk predictor.

Independent SNPs

Uncorrelated single-nucleotide polymorphisms (SNPs) in linkage equilibrium.

Linkage disequilibrium

(LD). The nonrandom association of alleles at different loci.

Polygenic prediction analysis

Any analysis method that predicts genetic risk or breeding values on the basis of the combined contribution of many loci.

Profile scoring

A polygenic prediction method for prediction of genetic value or risk for each individual (a 'profile') in a validation sample generated from the sum of the alleles they carry weighted by the association effect size estimated in a discovery sample.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Wray, N., Yang, J., Hayes, B. et al. Pitfalls of predicting complex traits from SNPs. Nat Rev Genet 14, 507–515 (2013).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing