1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
|
\name{roast.DGEList}
\alias{roast.DGEList}
\alias{mroast.DGEList}
\alias{fry.DGEList}
\title{Self-contained Gene Set Tests for Digital Gene Expression Data}
\description{
Rotation gene set testing for Negative Binomial generalized linear models.
}
\usage{
\method{fry}{DGEList}(y, index = NULL, design = NULL, contrast = ncol(design), geneid = NULL,
sort = "directional", ...)
\method{roast}{DGEList}(y, index = NULL, design = NULL, contrast = ncol(design), geneid = NULL,
set.statistic = "mean", gene.weights = NULL, nrot = 1999, \dots)
\method{mroast}{DGEList}(y, index = NULL, design = NULL, contrast = ncol(design), geneid = NULL,
set.statistic = "mean", gene.weights = NULL, nrot = 1999,
adjust.method = "BH", midp = TRUE, sort = "directional", \dots)
}
\arguments{
\item{y}{\code{DGEList} object.}
\item{index}{index vector specifying which rows (probes) of \code{y} are in the test set.
Can be a vector of integer indices, or a logical vector of length \code{nrow(y)}, or a vector of gene IDs corresponding to entries in \code{geneid}.
Alternatively it can be a data.frame with the first column containing the index vector and the second column containing directional gene weights.
For \code{mroast} or \code{fry}, \code{index} is a list of index vectors or a list of data.frames. }
\item{design}{the design matrix. Defaults to \code{y$design} or, failing that, to \code{model.matrix(~y$samples$group)}.}
\item{contrast}{contrast for which the test is required.
Can be an integer specifying a column of \code{design}, or the name of a column of \code{design}, or a numeric contrast vector of length equal to the number of columns of \code{design}.}
\item{geneid}{gene identifiers corresponding to the rows of \code{y}.
Can be either a vector of length \code{nrow(y)} or the name of the column of \code{y$genes} containing the gene identifiers.
Defaults to \code{rownames(y)}.}
\item{set.statistic}{summary set statistic. Possibilities are \code{"mean"},\code{"floormean"},\code{"mean50"} or \code{"msq"}.}
\item{gene.weights}{numeric vector of directional (positive or negative) genewise weights.
For \code{mroast} or \code{fry}, this vector must have length equal to \code{nrow(y)}.
For \code{roast}, can be of length \code{nrow(y)} or of length equal to the number of genes in the test set.}
\item{nrot}{number of rotations used to compute the p-values.}
\item{adjust.method}{method used to adjust the p-values for multiple testing. See \code{\link{p.adjust}} for possible values.}
\item{midp}{logical, should mid-p-values be used in instead of ordinary p-values when adjusting for multiple testing?}
\item{sort}{character, whether to sort output table by directional p-value (\code{"directional"}), non-directional p-value (\code{"mixed"}), or not at all (\code{"none"}).}
\item{\dots}{other arguments are currently ignored.}
}
\value{
\code{roast} produces an object of class \code{\link[limma:roast]{Roast}}. See \code{\link{roast}} for details.
\code{mroast} and \code{fry} produce a data.frame. See \code{\link{mroast}} for details.
}
\details{
These functions perform self-contained gene set tests against the null hypothesis that none of the genes in the set are differentially expressed.
\code{fry} is the recommended function in the edgeR context.
The roast gene set test was proposed by Wu et al (2010) for microarray data and the \code{roast} and \code{mroast} methods documented here extend the test to digital gene expression data.
The roast method uses residual space rotations instead of permutations to obtain p-values, a technique that take advantage of the full generality of linear models.
The negative binomial count data is converted to approximate normal deviates by computing mid-p quantile residuals (Dunn and Smyth, 1996; Routledge, 1994) under the null hypothesis that the contrast is zero, and the normal deviates are then passed to the limma \code{\link{roast}} function.
See \code{\link{roast}} for more description of the test and for a complete list of possible arguments.
\code{mroast} is similar but performs \code{roast} tests for multiple of gene sets instead of just one.
The \code{fry} method documented here similarly generalizes the fry gene set test for microarray data.
\code{fry} is recommended over \code{roast} or \code{mroast} for count data because, in this context, it is equivalent to \code{mroast} but with an infinite number of rotations.
}
\seealso{
\code{\link{roast}}, \code{\link{camera.DGEList}}
}
\author{Yunshun Chen and Gordon Smyth}
\references{
Dunn, PK, and Smyth, GK (1996).
Randomized quantile residuals.
\emph{J. Comput. Graph. Statist.}, 5, 236-244.
\url{https://gksmyth.github.io/pubs/residual.html}
Routledge, RD (1994).
Practicing safe statistics with the mid-p.
\emph{Canadian Journal of Statistics} 22, 103-110.
Wu, D, Lim, E, Francois Vaillant, F, Asselin-Labat, M-L, Visvader, JE, and Smyth, GK (2010). ROAST: rotation gene set tests for complex microarray experiments.
\emph{Bioinformatics} 26, 2176-2182.
\url{http://bioinformatics.oxfordjournals.org/content/26/17/2176}
}
\examples{
mu <- matrix(10, 100, 4)
group <- factor(c(0,0,1,1))
design <- model.matrix(~group)
# First set of 10 genes that are genuinely differentially expressed
iset1 <- 1:10
mu[iset1,3:4] <- mu[iset1,3:4]+10
# Second set of 10 genes are not DE
iset2 <- 11:20
# Generate counts and create a DGEList object
y <- matrix(rnbinom(100*4, mu=mu, size=10),100,4)
y <- DGEList(counts=y, group=group)
# Estimate dispersions
y <- estimateDisp(y, design)
roast(y, iset1, design, contrast=2)
mroast(y, iset1, design, contrast=2)
mroast(y, list(set1=iset1, set2=iset2), design, contrast=2)
}
\concept{Gene set testing}
|
