PGRN Network-Wide Project: Polygenic Modeling of Pharmacogenomic Traits

It is important for PGRN to have a set of analytic tools to evaluate available GWAS data for the contribution of “polygenic models” of pharmacological traits. Recently, two methods have been developed to detect the contribution of common SNPs in genome-wide association study (GWAS) data: polygenic analysis1 and mixed linear modeling2. Both methods test a polygenetic model in which many common SNPs in aggregate have a collective effect on phenotype.

In the first method, polygenic analysis, an additive polygenic risk score based on SNPs below a p-value threshold (PGWAS) in a discovery set of samples is then tested in an independent set of samples. Using this approach, polygenic effects have been demonstrated in schizophrenia1, multiple sclerosis3, height4, and body mass index (BMI)5. The second method, mixed linear modeling, estimates additive genetic variance under a mixed linear model with a random effect representing the polygenic component of trait variation. Applied to height2 and endometriosis6, this method demonstrated that common SNPs contribute to phenotypic variance.'

We hypothesize that there is a polygenic architecture underlying pharmacogenetic traits, and that many common variants in aggregate will predict response to therapy. Both of these methods will thus be applied to several of the unique datasets of the PGRN. First, we will use mixed linear modeling to determine the variance explained by common SNPs, within each phenotypic collection. Second, we will apply polygenic analysis as a complementary statistical approach. Both positive and negative results from these analyses are of interest, as results will inform us about the underlying genetic architecture of each trait and thus future discovery genetic studies. To help interpret results we will also conduct simulations based on GWAS sample size, trait definition, and different polygenic models of inheritance. We will also perform “power” and “prediction” analyses applying polygenic modeling results and systems biology “pathway” analyses, for phenotypes with evidence of a polygenic architecture.

  1. Purcell,S.M. et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature 460,748-52 (2009).
  2. Yang, J. et al. Common SNPs explain a large proportion of the heritability for human height. Nat Genet 42, 565-9 (2010).
  3. Bush, W.S. et al. Evidence for polygenic susceptibility to multiple sclerosis--the shape of things to come. Am J Hum Genet 86, 621-5 (2010).
  4. Lango Allen, H. et al. Hundres of variants clustered in genomic loci and biological pathways affect human height. Nature (2010).
  5. Speliotes, E.K. et al. Associaton analyses of 279,796 individuals reveal 18 new loci associated with body mass index. Nat Genet 42, 937-48 (2010).
  6. Painter, J.N. et al. Genome-wide association study identifies a locus at 7p15.2 associated with endometriosis. Nat Genet 43, 51-4 (2011).

Project Contacts

  • Robert Plenge, PI
  • Marylyn Ritchie, P-Star PI
  • Eli Stahl
  • Sarah Pendergrass

Participating PGRN groups:


  • PI, Admin/Management: Mary Relling
  • Analyst: Cheng Cheng


  • PI, Admin/Management: Alan Shuldiner
  • Analyst: Jeff O'Connell


  • PI, Admin/Management: Jerome Rotter
  • Analyst: Xiaohui Li


  • PI, Admin/Management: Julie Johnson
  • Analyst: Jeff O'Connell; Yan Gong


  • PI, Admin/Management: Kelan Tantisira
  • Analyst: Michael McGeachie; George Clemmer


  • PI, Admin/Management: Robert Plenge; Eli A. Stahl
  • Analyst: Jing Cui


  • PI, Admin/Management: Deanna Kroetz; John Witte
  • Analyst: Aparna Chhibber; Joel Mefford


  • PI, Admin/Management: Eric Gamazon; Nancy Cox
  • Analyst: TBD


A network resource for coordination of statistical analysis and methods development in the PGRN.


Polygenic Modeling