genpwr: Power Calculations Under Genetic Model Misspecification
Power and sample size calculations for genetic association studies allowing
for misspecification of the model of genetic susceptibility.
"Hum Hered. 2019;84(6):256-271.<doi:10.1159/000508558>. Epub 2020 Jul 28."
Power and/or sample size can be calculated for logistic (case/control study design)
and linear (continuous phenotype) regression models, using additive, dominant,
recessive or degree of freedom coding of the genetic covariate while assuming
a true dominant, recessive or additive genetic effect. In addition, power and
sample size calculations can be performed for gene by environment interactions.
These methods are extensions of Gauderman (2002)
<doi:10.1093/aje/155.5.478> and Gauderman (2002) <doi:10.1002/sim.973>
and are described in:
Moore CM, Jacobson S, Fingerlin TE. Power and Sample Size Calculations
for Genetic Association Studies in the Presence of Genetic Model Misspecification.
American Society of Human Genetics.
October 2018, San Diego.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=genpwr
to link to this page.