Computes expression measures for probe level data
express(x, method=c("avglogpm", "li.wong", "avdiff"), normalize=T, normalize.method=c("quantile", "loess", "contrasts"), span=2/3, choose.subset=T, subset.size=5000, verbose=T, maxit=1)
x |
a probe.level.object |
method |
which method to use to compute measure of expression, "avglogpm", "li.wong", or "avdiff". |
normalize |
logical value. If TRUE normalize probe level
data before computing expression measures. |
normalize.method |
which normalization method to use, "quantile", "loess" or "contrasts". |
span |
loess span to use with "loess" and "contrast"
methods |
choose.subset |
logical value. For the "loess" and "contrast"
methods, if TRUE it will do normalization on a subset and predict. |
subset.size |
size of subset to use in normalization. |
verbose |
logical value. If TRUE it writes out some messages. |
maxit |
maximum number of iterations to use in method "loess". |
"avglogpm" is the recommended method. It uses a background corrected average of PM values as a measure of expression. "li.wong" uses the Li and Wong algorithm. "avdiff" uses Genechip's average difference.
A matrix with expression values (genes represented by rows)
Rafael A. Irizarry
Irizarry, RA, Hobbs, B, and Speed, T (2001) http://biosun01.biostat.jhsph.edu/~ririzarr/papers/index.html