GDBR: GDBR: Genomic Distance-Based Regression

In gastonstat/AssotesteR: Statistical Tests for Genetic Association Studies

Description

The Genomic Distance-Based Regression has been developed by Wessel et al (2006). This approach captures genotype information across multiple loci through a similarity measure between any two individuals. GDBR is unique in its regression analysis relating variation in the measure of genomic similarity to variation in their trait values. Note that this approach is computationally expensive.

Usage

  GDBR(y, X, distance = "IBS", weights = NULL, perm = NULL)

Arguments

y	numeric vector with phenotype status: 0=controls, 1=cases. No missing data allowed
X	numeric matrix or data frame with genotype data coded as 0, 1, 2.
distance	character string indicating the type of distance to be used. Possible options are "IBS" or "wIBS" (distance="IBS" by default)
weights	optional numeric vector with weights for the genetic variants (NULL by default)
perm	positive integer indicating the number of permutations (NULL by default)

Details

The argument distance is used to specify the similarity distance. "IBS" indicates Identity-By-Share, "wIBS" indicates weighted IBS.

Value

An object of class "assoctest", basically a list with the following elements:

gdbr.stat	gdbr statistic
perm.pval	permuted p-value
args	descriptive information with number of controls, cases, variants, permutations, and selected distance
name	name of the statistic

Note

This method is computationally expensive

Author(s)

Gaston Sanchez

References

Wessel J, Schork NJ (2006) Generalized Genomic Distance-Based Regression Methodology for Multilocus Association Analysis. The American Journal of Human Genetics, 79: 792-806

Schaid DJ (2010) Genomic Similarity and Kernel Methods I: Advancements by Building on Mathematical and Statistical Foundations. The American Journal of Human Heredity, 70: 109-131

See Also

SKAT

Examples

  ## Not run: 
   
  # number of cases
  cases = 250

  # number of controls
  controls = 250

  # total (cases + controls)
  total = cases + controls

  # phenotype vector
  phenotype = c(rep(1,cases), rep(0,controls))

  # genotype matrix with 10 variants (random data)
  set.seed(123)
  genotype = matrix(rbinom(total*10, 2, 0.05), nrow=total, ncol=10)

  # apply GDBR with 50 permutations
  # (it takes some time to run the permutations!)
  mygdbr = GDBR(phenotype, genotype, perm=50)
  mygdbr
  
## End(Not run)

gastonstat/AssotesteR documentation built on May 16, 2019, 5:43 p.m.