Display options
Share it on

BMC Proc. 2014 Jun 17;8:S9. doi: 10.1186/1753-6561-8-S1-S9. eCollection 2014.

Evaluation of gene-based association tests for analyzing rare variants using Genetic Analysis Workshop 18 data.

BMC proceedings

Andriy Derkach, Jerry F Lawless, Daniele Merico, Andrew D Paterson, Lei Sun

Affiliations

  1. Department of Statistical Sciences, University of Toronto, Toronto, Ontario M5S 3G3, Canada.
  2. Department of Statistics and Actuarial Science, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada ; Division of Biostatistics, Dalla Lana School of Public Health, University of Toronto, Ontario M5S 3G3, Canada.
  3. The Centre for Applied Genomics, The Hospital for Sick Children, Toronto, Ontario M5G 1L7, Canada ; Program in Genetics and Genome Biology, The Hospital for Sick Children, Toronto M5G 1X8, Canada.
  4. Division of Biostatistics, Dalla Lana School of Public Health, University of Toronto, Ontario M5S 3G3, Canada ; Program in Genetics and Genome Biology, The Hospital for Sick Children, Toronto M5G 1X8, Canada.
  5. Department of Statistical Sciences, University of Toronto, Toronto, Ontario M5S 3G3, Canada ; Division of Biostatistics, Dalla Lana School of Public Health, University of Toronto, Ontario M5S 3G3, Canada.

PMID: 25519417 PMCID: PMC4143759 DOI: 10.1186/1753-6561-8-S1-S9

Abstract

The focus of our work is to evaluate several recently developed pooled association tests for rare variants and assess the impact of different gene annotation methods and binning strategies on the analyses of rare variants under Genetic Analysis Workshop 18 real and simulated data settings. We considered the sample of 103 unrelated individuals with sequence data, genotypes of rare variants from chromosome 3, real phenotype of hypertension status and simulated phenotypes of systolic blood pressure (SBP) and diastolic blood pressure (DBP), and covariates of age, sex, and the interaction between age and sex. In the analysis of real phenotype data, we did not obtain significant results for any binning strategy; however, we observed a slight deviation of the p-values from the uniform distribution based on the protein-damaging variant grouping strategy. Evaluation of methods using simulated data showed lack of power even at the conservative level of 0.05 for most of the causal genes on chromosome 3. Nevertheless, analysis of MAP4 produced good power for all tests at various levels of the tests for both DBP and SBP. Our results also confirmed that Fisher's method is not only robust but can also improve power over individual pooled linear and quadratic tests and is often better than other robust tests such as SKAT-O.

References

  1. Biostatistics. 2012 Sep;13(4):762-75 - PubMed
  2. Mutat Res. 2007 Feb 3;615(1-2):28-56 - PubMed
  3. Am J Hum Genet. 2011 Sep 9;89(3):354-67 - PubMed
  4. PLoS Genet. 2009 Feb;5(2):e1000384 - PubMed
  5. Genome Res. 2005 Aug;15(8):1034-50 - PubMed
  6. BMC Proc. 2014 Jun 17;8(Suppl 1 Genetic Analysis Workshop 18Vanessa Olmo):S11 - PubMed
  7. Nat Methods. 2010 Apr;7(4):248-9 - PubMed
  8. PLoS Genet. 2011 Mar;7(3):e1001322 - PubMed
  9. Am J Hum Genet. 2011 Apr 8;88(4):440-9 - PubMed
  10. Nat Protoc. 2009;4(7):1073-81 - PubMed
  11. Genet Epidemiol. 2013 Jan;37(1):110-21 - PubMed
  12. Nucleic Acids Res. 2010 Sep;38(16):e164 - PubMed

Publication Types

Grant support