IsoRawp {IsoGene} | R Documentation |
The function calculates the raw one-sided and two-sided p-values for each test statistic using permutations.
IsoRawp(x, y, niter)
x |
numeric vector containing the dose levels |
y |
a data frame of the gene expression with Probe IDs as row names |
niter |
number of permutations to use |
The number of permutations to use can be chosen based on the number of possible permutations of samples. If the possible number is too big, usually >5000 permutations can be sufficient.
A list of components
raw.p.one |
returns the one-sided p-value matrix for the five test statisticsin in 6 columns: the first column is the probe ID, the second to the last columns contain the raw p-values for each test statistic |
raw.p.two |
returns the two-sided p-value matrix for the five test statistics in 6 columns: the first column is the probe ID, the second to the last columns contain the raw p-values for each test statistic |
rawp.up |
returns the one-sided p-value matrix testing increasing alternative for the five test statistics in 6 columns: the first column is the probe ID, the second to the last columns contain the raw p-values for each test statistic |
rawp.dn |
returns the one-sided p-value matrix testing decreasing alternative for the five test statistics in 6 columns: the first column is the probe ID, the second to the last columns contain the raw p-values for each test statistic |
For each gene, the one-sided p-values are calculated from min(p^Up, p^Down) and the two sided p-values are calculated from min{2 * min(p^Up, p^Down), 1}, where p^Up and p^Down are the p-values calculated for each ordered alternative.
Lin et al.
Lin et al. (2007). Microarray Experiments: a Comparis. on of Testing Procedures, Multiplicity, and Resampling-Based Inference, Stat. App. in Gen. & Mol. Bio., 6(1), article 26.
## Not run: set.seed(1234) x <- c(rep(1,3),rep(2,3),rep(3,3)) y1 <- matrix(rnorm(90, 1,1),10,9) # 10 genes with no trends y2 <- matrix(c(rnorm(30, 1,1), rnorm(30,2,1), rnorm(30,3,1)), 10, 9) # 10 genes with increasing trends y <- data.frame(rbind(y1, y2)) # y needs to be a data frame rp <- IsoRawp(x, y, niter = 1000) rp ## End(Not run)