comp.stat: Computing Test Statistics for Differential Expression

In DEDS: Differential Expression via Distance Summary for Microarray Data

Description

This function computes test statistics, e.g., t-statistics, F-statistics, SAM, fold changes, moderated t or F statistics, B statistics, for each row of a microarray data matrix.

Usage

comp.stat(X, L, test = c("t", "fc", "sam", "f", "modt", "modf", "B"), extra = NULL)

Arguments

X	A matrix, with m rows corresponding to variables (hypotheses) and n columns to observations. In the case of gene expression data, rows correspond to genes and columns to mRNA samples. The data can be read using read.table.
L	A vector of integers corresponding to observation (column) class labels. For k classes, the labels must be integers between 0 and k-1.
test	A character string specifying the statistic to be used to test the null hypothesis of no association between the variables and the class labels. test="t": t-statistics; test="f": F-statistics; test="fc": fold changes; test="sam": SAM-statistics; test="modt": moderated t-statistics; test="modf": moderated F-statistics; test="B": B-statistics.
extra	Extra parameter needed for the test specified; see deds.genExtra.

Details

The function comp.stat interfaces to a C function and computes various statistics for differential expression in the C environment and therefore faster than functions in R. However, functions in R that are implemented in the DEDS packages may have more flexibility in terms of specifications of arguments. Below is a table the details comp.stat and its equivalent R functions in the DEDS package. Note that all the R functions listed in the 2nd column of the table below return a function with bindings for a series of arguments which accept the microarray data matrix as its single argument and compute accordingly statistics.

Interface to C	R functions	Statistics
deds.stat(X, L, test="t")	tTest(L=NULL, mu=0, var.equal=FALSE)	t statistics
deds.stat(X, L, test="fc")	FC(L=NULL, is.log=TRUE, FUN=mean)	fold change
deds.stat(X, L, test="sam")	Sam(L=NULL, prob=0.5, B=200, stat.only=TRUE, verbose=FALSE, deltas, s.step=0.01, alpha.step=0.01, plot.it=FALSE)	SAM statistics
deds.stat(X, L, test="f")	fTest(L=NULL)	F statistics
deds.stat(X, L, test="modt")	tmodTest(L=NULL)	moderated t statistics
deds.stat(X, L, test="modf")	fmodTest(L=NULL)	moderated F statistics
deds.stat(X, L, test="B")	BTest(L=NULL, proportion=0.01)	B statistics

Value

A vector of test statistics for each row of the matrix.

Author(s)

Yuanyuan Xiao, [e-mail],
Jean Yee Hwa Yang, [e-mail].

References

For references on B-statistics and moderated t and F statistics:

Lonnstedt, I. and Speed, T. P. (2002). Replicated microarray data. Statistica Sinica 12, 31-46.

Smyth, G. K. (2003). Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. http://www.statsci.org/smyth/pubs/ebayes.pdf

See Also

deds.genExtra, for B statistics: lm.series and ebayes

Examples

X <- matrix(rnorm(1000,0,0.5), nc=10)
L <- rep(0:1,c(5,5))

# genes 1-10 are differentially expressed
X[1:10,6:10]<-X[1:10,6:10]+1

# t statistics
tstat <- comp.stat(X, L, test="t")

# SAM, fudge factor set as the median of pooled genewise standard deviations
samstat <- comp.stat(X, L, test="sam")
# SAM, fudge factor set as the 90% of pooled genewise standard deviations
samstat <- comp.stat(X, L, test="sam", extra=c(0.9))

# moderated t
modtstat <- comp.stat(X, L, test="modt")

# B, proportion of differentially expressed genes is set at default, 1%
Bstat <- comp.stat(X, L, test="B")
# B, proportion of differentially expressed genes is set at  10%
Bstat <- comp.stat(X, L, test="B", extra=c(0.1))

DEDS documentation built on Oct. 31, 2019, 3:15 a.m.