normal quantile transformation

Question

I need to do quantile normalization in R

Input file just a small subset I have over 5000 values:

x <- data.frame(A =c(193973, 185750, 185511,NA), B= c(56433,52298, 53040, NA), C = c(4668, 6074,6246, NA))

software I am using is recommending

xtransformed = qqnorm(x, plot.it =F)$x

I don't understand "$x" here in the end ?

other question is:

if input is (which is my actual input

x <- data.frame(A =c("A","B","C","D"), B = c("X", "Y", "Z","L"), C=c(193973, 185750, 185511, "NA"), D = c(56433,52298, 53040, "NA"), E = c(4668, 6074,6246, "NA"))

and I just select column I want to transform

y =x[,3:5]

Transformed it with that code and it doesnt work? What is the reason

ytransformed <- qqnorm(y, plot.it =F)$y

ytransformed 

I did that. Now I want to plot and see if it has transformed or not? how can I do that?

Answer 1

The preprocessCore package in Bioconductor has this function. First install & load it:

source('http://bioconductor.org/biocLite.R')
biocLite('preprocessCore')
library(preprocessCore)

Then make a matrix out of your data frame, and normalize.quantiles :

> x <- data.frame(A =c(193973, 185750, 185511,NA), B= c(56433,52298, 53040, NA), C = c(4668, 6074,6246, NA))
> m <- data.matrix(x,rownames.force=nrow(x))
> normalize.quantiles(m)
         [,1]     [,2]     [,3]
[1,] 85550.67 85550.67 80825.67
[2,] 82143.61 80825.67 82143.61
[3,] 80825.67 82143.61 85550.67
[4,]       NA       NA       NA