| Title: | Gene signature generation for functionally validated signaling pathways |
|---|---|
| Description: | By leveraging statistical properties (log-rank test for survival) of patient cohorts defined by binary thresholds, poor-prognosis patients are identified by the sigsquared package via optimization over a cost function reducing type I and II error. |
| Authors: | UnJin Lee |
| Maintainer: | UnJin Lee <[email protected]> |
| License: | GPL version 3 |
| Version: | 1.37.0 |
| Built: | 2024-06-30 04:56:36 UTC |
| Source: | https://github.com/bioc/sigsquared |
The analysisPipeline function is used to train a set of thresholds for predicting survival outcome within the context of a given signaling environment. This signaling environment is encoded in a geneSignature object.
analysisPipeline(dataSet, geneSig, iterPerK=2500, k=3, rand=TRUE, newjpdf=FALSE, jpdf=FALSE, nJPDF=12500, disc=c(0.005, 0.01, 0.03, 0.05), MFS="MFS", met="met", optMeth="Nelder-Mead")analysisPipeline(dataSet, geneSig, iterPerK=2500, k=3, rand=TRUE, newjpdf=FALSE, jpdf=FALSE, nJPDF=12500, disc=c(0.005, 0.01, 0.03, 0.05), MFS="MFS", met="met", optMeth="Nelder-Mead")
dataSet |
ExpressionSet object containing both expression data (exprs) and phenotypic survival data (pData) |
geneSig |
geneSignature object containing directions, thresholds, and gene symbols |
iterPerK |
integer number of optimization iterations for each k |
k |
integer k for k-fold cross-validation |
rand |
boolean determining whether the k subsets are randomly drawn (otherwise k subsets are selected ordinally) |
newjpdf |
boolean for generating a joint probability function for alternate smoothed cost function (not recommended) |
jpdf |
solnSpace object containing empirical joint probability function for alternate smoothed cost function (not recommended) |
nJPDF |
value determining the number of samples with which to estimate the empirical joint probability function for alternate smoothed cost function (not recommended) |
disc |
vector of discretation thresholds for discretized cost function |
MFS |
variable name for survival-time data in dataSet object |
met |
variable name for metastasis event data in dataSet object |
optMeth |
optimization method used by R function 'optim' |
The analysisPipeline function optimizes over a cost function designed to minize both type I and II error. There is a discretized and smoothed cost function available, however implementation of the smoothed cost function relies on sampling of the solution space. This sampling may be pre-computed and implemented through the 'jpdf' argument, however overall usage of the smoothed cost function is not recommended.
A geneSignature object containing newly trained thresholds
UnJin Lee
## Load in example data data("BrCa443") ## Create initial geneSignature object ## Note it is not necessary to define thresholds at this point gs <- setGeneSignature(g=new("geneSignature"), direct=c(-1,1,1,1,1,1,1), genes=c("RKIP", "HMGA2", "SPP1", "CXCR4", "MMP1", "MetaLET7", "MetaBACH1")) ## Generate thresholds gs <- analysisPipeline(dataSet=BrCa443, geneSig=gs, iterPerK=50, k=2, rand=FALSE)## Load in example data data("BrCa443") ## Create initial geneSignature object ## Note it is not necessary to define thresholds at this point gs <- setGeneSignature(g=new("geneSignature"), direct=c(-1,1,1,1,1,1,1), genes=c("RKIP", "HMGA2", "SPP1", "CXCR4", "MMP1", "MetaLET7", "MetaBACH1")) ## Generate thresholds gs <- analysisPipeline(dataSet=BrCa443, geneSig=gs, iterPerK=50, k=2, rand=FALSE)
BrCa443 is an ExpressionSet object that contains gene expression values for 7 genes, RKIP, HMGA2, SPP1, CXCR4, MMP1, metaBACH1, and metaLET7 for 443 breast cancer patients. It also contains paired survival data for each patient in the form of survival and event data.
data(BrCa443)data(BrCa443)
ExpressionSet with expression data and survival data
GSE5327, GSE2034, and GSE2603
The ensembleAdjustable function applies a geneSignature object to a data matrix containing expression values and gene symbols or an ExpressionSet object.
ensembleAdjustable(dataSet, geneSig, index=F)ensembleAdjustable(dataSet, geneSig, index=F)
dataSet |
data set object, may be numeric matrix or an ExpressionSet |
geneSig |
geneSignature object containing directions, thresholds, and gene symbols |
index |
index to indicate which samples are to be subsetted, may be FALSE for no subsetting or a vector of column numbers |
A logical vector with length equal to the number of samples (or samples subsetted), TRUE indicating a positive, FALSE indicating a negative
UnJin Lee
require(Biobase) ## Generate test geneSignature object with 0s for thresholds gs <- setGeneSignature(g=new("geneSignature"), direct=c(1,1,1), genes=c("A", "B", "C"), thresholds=c(0, 0, 0)) ## Generate randomly distributed matrix and ExpressionSet mat <- matrix(rnorm(9, 0, 1), nrow=3) rownames(mat) <- c("A", "B", "C") posmat <- abs(mat) expset <- new("ExpressionSet", exprs=mat) ## Apply geneSignature to matrices ensembleAdjustable(mat, gs) ensembleAdjustable(posmat, gs) ## Apply geneSignature to ExpressionSet ensembleAdjustable(expset, gs) ## Apply geneSignature with subsetting ensembleAdjustable(mat, gs, c(1, 3)) ensembleAdjustable(expset, gs, c(1, 3))require(Biobase) ## Generate test geneSignature object with 0s for thresholds gs <- setGeneSignature(g=new("geneSignature"), direct=c(1,1,1), genes=c("A", "B", "C"), thresholds=c(0, 0, 0)) ## Generate randomly distributed matrix and ExpressionSet mat <- matrix(rnorm(9, 0, 1), nrow=3) rownames(mat) <- c("A", "B", "C") posmat <- abs(mat) expset <- new("ExpressionSet", exprs=mat) ## Apply geneSignature to matrices ensembleAdjustable(mat, gs) ensembleAdjustable(posmat, gs) ## Apply geneSignature to ExpressionSet ensembleAdjustable(expset, gs) ## Apply geneSignature with subsetting ensembleAdjustable(mat, gs, c(1, 3)) ensembleAdjustable(expset, gs, c(1, 3))
"geneSignature"
The geneSignature object contains the necessary elements defining the signaling environment on which a prognostic gene signature will be created.
Objects can be created by calls of the form new("geneSignature", ...).
Objects all contain 4 slots - geneSet, geneDirect, thresholds, dirMat (unused).
geneSet:Object of class "character" ~~
geneDirect:Object of class "numeric" ~~
thresholds:Object of class "numeric" ~~
dirMat:Object of class "matrix" ~~
signature(dataSet = "ExpressionSet", geneSig = "geneSignature"): ...
signature(dataSet = "ExpressionSet", geneSig = "geneSignature"): ...
signature(dataSet = "matrix", geneSig = "geneSignature"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature", direct = "numeric"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature"): ...
signature(g = "geneSignature"): ...
UnJin lee
Lee U, Frankenberger C, Yun J, Bevilacqua E, Caldas C, et al. (2013) A Prognostic Gene Signature for Metastasis-Free Survival of Triple Negative Breast Cancer Patients. PLoS ONE 8(12): e82125. doi:10.1371/journal.pone.0082125
showClass("geneSignature")showClass("geneSignature")
The geneSignature object contains the necessary elements defining the signaling environment on which a prognostic gene signature will be created. This collection of functions are used to manipulate or retrieve the data slots of a given geneSignature object.
setGeneSignature(g, direct=NA, thresholds=c(0), genes=NA, mat=matrix()) setDirect(g, direct) setThresholds(g, thresholds) setGenes(g, genes) getDirect(g) getThresholds(g) getGenes(g) getNGenes(g)setGeneSignature(g, direct=NA, thresholds=c(0), genes=NA, mat=matrix()) setDirect(g, direct) setThresholds(g, thresholds) setGenes(g, genes) getDirect(g) getThresholds(g) getGenes(g) getNGenes(g)
g |
geneSignature object |
direct |
vector of -1s or 1s representing down- or up-regulation respectively |
thresholds |
vector of values containing thresholds for the geneSignature object |
genes |
character vector of gene names |
mat |
matrix of interactions between genes (unused) |
All setting functions return objects of class geneSignature. getDirect yields a vector of -1s or 1s, getThesholds yields a vector of theshold values, getGenes yields a character vector of gene names, getNGenes yields the number of genes in the geneSignature
UnJin Lee
## Generate and read out values of a geneSignature object gs <- setGeneSignature(new("geneSignature"), c(1, 1), c(0, 0), c("BACH1", "RKIP"), matrix()) getDirect(gs) getThresholds(gs) getGenes(gs) getNGenes(gs)## Generate and read out values of a geneSignature object gs <- setGeneSignature(new("geneSignature"), c(1, 1), c(0, 0), c("BACH1", "RKIP"), matrix()) getDirect(gs) getThresholds(gs) getGenes(gs) getNGenes(gs)
The sigsquared package attempts to detect the presence of alternate signaling states of an input pathway that significantly predict differential survival outcome within a mixed cohort of patients. The main goal of this package is to generate gene signatures given known signaling pathways that are predictive of differential survival outcome. The two main functions used to accomplish this goal are first, the ability to train model parameters for a given linear network model, and second, the ability to apply the model and trained parameters to transcript data.