| data {emBayes} | R Documentation |
simulated gene expression example data
Description
Simulated gene expression data for demonstrating the usage of emBayes.
Usage
data(data)
Format
The data file consists of five components: y, clin, X, quant, coef and clin.coe. The coefficients and clinical coefficients are the true values of parameters used for generating response y. They can be used for performance evaluation.
Details
The data model for generating response
Let y_{i} be the response of the i-th subject (1\leq i\leq n). We have z_{i}=(1,z_{i1},\dots,z_{iq})^{\top} being a (q+1)-dimensional vector of which the last q components indicate clinical factors and x_{i}=(x_{i1},\dots,x_{ip})^{\top} denoting a p-dimensional vector of genetic factors. The linear quantile regression model for the \tau-th quantile (0<\tau<1) is:
y_i=z_i^\top\alpha+x_i^\top\beta+\epsilon_i
where \alpha=(\alpha_0,\cdots,\alpha_q)^\top contains the intercept and the regression coefficients for the clinical covariates. \beta=(\beta_1,\cdots,\beta_p)^\top are the regression coefficients and random error \epsilon_{i}=(\epsilon_{1},...,\epsilon_{n})^\top is set to follow a T2 distribution and has value 0 at its \tau-th quantile.