jung.rstat: Function to Compute Jung's Statistics
In PROMISE: PRojection Onto the Most Interesting Statistical Evidence

Description Usage Arguments Value Note Author(s) References See Also Examples

Compute statistic that measures the correlation of many continuous variables with a censored time-to-event variable

1	jung.rstat(x, time.cens, strat = NULL)

`x`	a data frame with row corresponding to probe set and column corresponding to subjects, the order of columns (subjects) should match the order of rows in time.cens.
`time.cens`	a data frame with number of row equal to number of column in x. It contains two columns with first for time and second for censor (1 = event, 0 = censored).
`strat`	a vector of stratum to calculate stratified r-type association statistics, default = NULL.

Returns a vector of Jun's r-type association statistics.

The order of subjects in x (column), time.cens, and strat should all match. The original statistic proposed by Jung, Owzar, and George can be written as a dot-product. The statistic returned by this routine is expressed in the form of a correlation statistic by dividing the dot product by the square root of the lengths of the two vectors in the numerator.

Stan Pounds stanley.pounds@stjude.org; Xueyuan Cao xueyuan.cao@stjude.org

Jung SH, Owzar K, and George SL (2005) A multiple testing procedure to associate gene expression levels with survival. Stat Med 24:3077-88

PROMISE

## load sampExprSet.
data(sampExprSet)

## extract expression matrix from sampExprSet
Y <- exprs(sampExprSet)

## extract end point data from sampExprSet
time.cens <-pData(phenoData(sampExprSet))[, 3:4] 
strat <- pData(phenoData(sampExprSet))$strat

## compute Jung's r-type association statistics
jungstat <- jung.rstat(Y, time.cens, strat = strat)