Adjust for cell composition in 450K data.

Much of the adjust.beta code was originally written by Andres Houseman. I have only put it into a simple to use function and packaged it.

Installation

> library(devtools)
> install_github("brentp/celltypes450")

About

See this thread for details.

Compared to minfi, which normalizes the sorted data with the new data:

To get your beta (0, 1) values adjusted for cell-type composition, use:

library(devtools)
install_github("brentp/celltypes450")
library(celltypes450)
adjusted = adjust.beta(beta)

To get the cell-composition estimates such as those from the image above:

cell.types = adjust.beta(beta, est.only=TRUE)