Description Usage Arguments Details Value Warning References Examples
View source: R/WEE_functions.r View source: R/WEE_functions.r
Returns an object of class "WEE.quantile" that is generated by quantile regression with WEE approach for continuous secondary traits in genetic case-control studies.
1 | WEE.quantile(formula, D, data, pd_pop, tau, iter = 5, boot = 0, ...)
|
formula |
The secondary trait given SNPs and covariates. e.g. y~x+z |
D |
Primary disease (case-control status), must be specified. |
data |
Dataset with real observation. |
pd_pop |
The population disease prevelance of primary disease. |
tau |
The quantile level to be estimated. Multiple taus can be chosen. |
iter |
Number of generating pseudo observations. (iter=10 by default) |
boot |
Number of bootstrape samples. (boot=0 by default) |
... |
Optional arguments to be passed through to rq. |
The quantile regression package "quantreg" is required before calling this function
Coefficients |
Point estimates |
StdErr |
Bootstrap standard errors, returned if boot > 0 |
Wald |
Wald test statistics, returned if boot > 0 |
p.value |
p-values, returned if boot > 0 |
Covariance |
Covariance matrix, returned if boot > 0 |
If boot = 0, point estimates are plotted. If boot > 0, boostrap standard errors, Wald test statistics, p-values, and covariance matrix are also returned. Optional arguments from rq can be passed to this function, but arguments 'subset' and 'weights' should be used with caution.
[1] Ying Wei, Xiaoyu Song, Mengling Liu, Iuliana Ionita-Laza and Joan Reibman (2016). Quantile Regression in the Secondary Analysis of Case Control Data. Journal of the American Statistical Association, 111:513, 344-354; DOI: 10.1080/01621459.2015.1008101
[2] Xiaoyu Song, Iuliana Ionita-Laza, Mengling Liu, Joan Reibman, Ying Wei (2016). A General and Robust Framework for Secondary Traits Analysis. Genetics, vol. 202 no. 4 1329-1343; DOI: 10.1534/genetics.115.181073
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 | ## Generate simulated data
# set population size as 500000
n = 500000
# set parameters
beta = c(0.12, 0.1) # P(Y|X,Z)
gamma = c(-4, log(1.5), log(1.5), log(2) ) #P(D|X,Y,Z)
# generate the genetic variant X
x = rbinom(n,size=2,prob=0.3)
# generate the continuous covariate Z
z = rnorm(n)
# generate the continuous secondary trait Y
y= 1 + beta[1]*x + beta[2]*z + (1+0.02*x)*rnorm(n)
# generate disease status D
p = exp(gamma[1]+x*gamma[2]+z*gamma[3]+y*gamma[4])/
(1+exp(gamma[1]+x*gamma[2]+z*gamma[3]+y*gamma[4]))
d = rbinom(n,size=1,prob=p)
# form population data dataset
dat = as.data.frame(cbind(x,y,z,d))
colnames(dat) = c("x","y","z","d")
# Generate sample dataset with 200 cases and 200 controls
dat_cases = dat[which(dat$d==1),]
dat_controls= dat[which(dat$d==0),]
dat_cases_sample = dat_cases[sample(sum(dat$d==1),
200, replace=FALSE),]
dat_controls_sample = dat_controls[sample(sum(dat$d==0),
200, replace=FALSE),]
dat_quantile = as.data.frame(rbind(dat_cases_sample,
dat_controls_sample))
colnames(dat_quantile) = c("x","y","z","D")
D = dat_quantile$D # Disease status
pd = sum(d==1)/n # population disease prevalence
# WEE quantile regressions:
WEE.quantile(y ~ x, D, tau = 0.5,
data = dat_quantile, pd_pop = pd)
WEE.quantile(y ~ x + z, D, tau = 1:9/10,
data = dat_quantile, pd_pop = pd, boot = 500)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.