Description Usage Arguments Details Value Author(s) References See Also Examples
High-level function for package SigCheck
that runs a default
set of checks against a predictive signature.
1 2 3 | sigCheckAll(check,
iterations=10, known="cancer",
plotResults=TRUE, ...)
|
check |
A |
iterations |
Number of iterations to run to generate background distributions. This is how many random signatures the primary signature will be compared to, as well as how many of each type of permuted dataset will be generated for comparison. |
known |
Specification of a set of known (previously identified) signatures to
compare to. See |
plotResults |
By default, plots of the results will be generated
unless this is set or |
... |
Extra parameters to pass through |
This high-level function will run four checks, plot the results, and return a consolidated result set.
First, it calls sigCheckRandom
to compare the performance
of interations
randomly selected signatures.
Next, it calls sigCheckKnown
to check the
performance of the signature against a database of signatures previously
identified to discriminate in other domains.
Finally, two calls are made to sigCheckPermuted
to check the
performance of randomly permuted metadata and expression data.
The first call permutes the survival data if they
are available (toPermute="survival"
);
otherwise it permutes the
category assignments (toPermute="categories"
)
The second call permuted the expression value for each gene (permuting
each row in the ExpressionSet
,
equivalent to toPermute="features"
).
If plotResults
is TRUE
, the results are plotted. If a classifier
is involved, a set of four classification results are plotted in
a 2x2 grid, showing how the classification performance of the main
signature compares to that of a mode classifier and to the
distribution of performance values observed for
the random and known signature sets, as well as how it performs using the two
type of permuted dataset.
If survival data is available, another 2x2 grid is plotted showing how the
baseline survival p-value compares to a p-value of 0.05 and to the
distribution of p-values observed for the random and known signatures, as well
as for the permuted data.
A list containing four elements, each containing the result of a check.
$checkRandom
is the result list returned by
sigCheckRandom
.
$checkKnown
is the result list returned by
sigCheckKnown
.
The third element of the result list will be one of the following:
$checkPermutedSurvival
is the result list returned by
sigCheckPermuted
with toPermute="survival"
.
$checkPermutedCategories
is the result list returned by
sigCheckPermuted
with toPermute="categories"
.
The fourth element of the list will be:
$checkPermutedFeatures
is the result list returned by
sigCheckPermuted
with toPermute="features"
.
Rory Stark
Venet, David, Jacques E. Dumont, and Vincent Detours. "Most random gene expression signatures are significantly associated with breast cancer outcome." PLoS Computational Biology 7.10 (2011): e1002240.
sigCheck
, sigCheckRandom
,
sigCheckPermuted
, sigCheckKnown
,
sigCheckPlot
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 | #Disable parallel so Bioconductor build won't hang
library(BiocParallel)
register(SerialParam())
library(breastCancerNKI)
data(nki)
nki <- nki[,!is.na(nki$e.dmfs)]
data(knownSignatures)
ITERATIONS <- 5 # should be at least 20, 1000 for real checks
## survival analysis
check <- sigCheck(nki, classes="e.dmfs", survival="t.dmfs",
signature=knownSignatures$cancer$VANTVEER,
annotation="HUGO.gene.symbol",
validationSamples=150:319)
results <- sigCheckAll(check,iterations=ITERATIONS,
known=knownSignatures$cancer[1:20])
## classification analysis
check <- sigCheck(nki, classes="e.dmfs",
signature=knownSignatures$cancer$VANTVEER,
annotation="HUGO.gene.symbol",
validationSamples=275:319,
scoreMethod="classifier")
results <- sigCheckAll(check,iterations=ITERATIONS,
known=knownSignatures$cancer[1:20])
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.