PROMISE: PRojection onto the Most Interesting Statistical Evidence

Description Usage Arguments Value Author(s) References See Also Examples

Description

Perform permutation-based test to identify genes with expression levels having a specific biologically interesting pattern of associations with multiple endpoint variables

Usage

1
2
PROMISE(exprSet, geneSet=NULL, promise.pattern, strat.var=NULL, proj0=FALSE, seed=13, nbperm=FALSE, max.ntail=100,
        nperms=10000)

Arguments

exprSet

an ExpressionSet class contains minimum of exprs (expression matrix) and phenoData (AnnotatedDataFrame of end point data). Please refer to Biobase for details on how to create such an ExpressionSet.

geneSet

a GeneSetCollection class with minimum of setName and geneIDs for each GeneSet. Please refer to GSEABase for how to create such a GeneSetCollection class. The default is NULL which will perform no gene set enrichment analysis.

promise.pattern

a data frame defining the association pattern of interest. The column names must be stat.coef, stat.func, and endpt.vars. The stat.coef column gives the coefficients for combining the statistics of association of genomic variables with individual endpoint variable into the ultimate PROMISE statistic. If proj0=TRUE, the stat.coeff is ignored. The stat.func column gives the name of the R routine that computes the test statistic of association of the endpoint variables. Two R routines (jung.rstat and spearman.rstat)are provided. Users can provide their own routine accordingly. The endpt.vars column gives the name(s) of variable(s) in the endpoint data file needed to compute each term of the PROMISE statistic. A common without a space should be used to separate multiple variables that correspond to the same term in the association pattern definition.

strat.var

the name or numeric value of stratum variable in exprSet for stratified analysis. The default is NULL which performs an unstratified analysis.

proj0

indicator of whether projection to 0 is performed. It takes two valid values: TRUE or FALSE. If proj0=TRUE, PROMISE statistics is the sum of squares of individual statitics and the stat.coeff in promise.pattern is ignored. The default is FALSE.

seed

initial seed of random number generator. The default is 13.

nbperm

indicator of fast permuation using negative binomial strategy, taking two valid values: FALSE or TRUE. The default is FALSE.

max.ntail

number of sucess if nbperm = T. Further permutation will not be performed for gene(s) or gene set(s) which max.ntail permutated statistics are greater or equal to the observed statistics, The default is 100

nperms

number of permutations. The default is 10,000.

Value

$generes

individual genes' test statistics and p-values for each individual endpoint and PROMISE analysis. If nbperm=T, the last column contains number of permuations for each gene.

$setres

gene set's test statistics and p-values for each individual endpoint and PROMISE analysis. If nbperm=T, the last column contains number of permuations for gene set. If geneSet is NULL, the value of this component is also NULL.

Author(s)

Stan Pounds stanley.pounds@stjude.org; Xueyuan Cao xueyuan.cao@stjude.org

References

Pounds S, Cheng C, Cao X, Crews KR, Plunkett W, Gandhi V, Rubnitz J, Ribeiro RC, Downing JR, and Lamba J (2009) PROMISE: a tool to identify genomic features with a specific biologically interesting pattern of associations with multiple endpoint variables. Bioinformatics 25: 2013-2019

See Also

jung.rstat avg.abs.genestat promise.genestat spearman.rstat promise.pattern

Examples

 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
## load sampExprSet, sampGeneSet, phPatt.
data(sampExprSet)
data(sampGeneSet)
data(phPatt)

## Perform PROMISE procedure without GSEA
test1 <- PROMISE(exprSet=sampExprSet,
                 geneSet=NULL, 
                 promise.pattern=phPatt, 
                 strat.var=NULL,
                 proj0=FALSE,
                 nbperm=FALSE, 
                 max.ntail=10,
                 seed=13, 
                 nperms=100)
             
## Perform PROMISE procedure with GSEA and using fast permuation              
test2 <- PROMISE(exprSet=sampExprSet,
                 geneSet=sampGeneSet, 
                 promise.pattern=phPatt, 
                 strat.var=NULL,
                 proj0=TRUE,
                 nbperm=TRUE, 
                 max.ntail=10,
                 seed=13, 
                 nperms=100)

PROMISE documentation built on Nov. 8, 2020, 5:15 p.m.