TSPC Manual
User Manual:
Open the PDF directly: View PDF 
.
Page Count: 9
Package ‘TSPC’
July 6, 2015
Type Package
Title Prediction using time-course gene expression
Version 2.0
Date 2014-01-17
Author Yuping Zhang
Maintainer Yuping Zhang <yuping.zhang@uconn.edu>
Description Performs survival and quantitative outcome using
time-course gene expression, described in the following
papers:
Zhang Y, Tibshirani RJ, Davis RW. Predicting patient survival from longitudinal gene expres-
sion. Stat Appl Genet Mol Biol. 2010;9(1):Article41. Epub 2010 Nov 22.
Zhang Y, Ouyang Z. Predicting quantitative outcomes of patients using longitudinal gene expres-
sion. Sri Lankan Journal of Applied Statistics, 5(4), 117-126.
License GPL (>= 2)
LazyLoad TRUE
Depends superpc, survival
Repository CRAN
Date/Publication 2012-10-29 08:57:45
Rtopics documented:
TSPC-package ....................................... 2
tspc.cv............................................ 2
tspc.plotcv.......................................... 4
tspc.predict ......................................... 5
tspc.project ......................................... 6
tspc.train........................................... 7
Index 9
1
2tspc.cv
TSPC-package TSPC
Description
Performs survival and quantitative outcome using time-course gene expression, described in the
following papers: Zhang Y, Tibshirani RJ, Davis RW. Predicting patient survival from longitudinal
gene expression. Stat Appl Genet Mol Biol. 2010;9(1):Article41. Epub 2010 Nov 22.\ Zhang Y,
Ouyang Z. Predicting quantitative outcomes of patients using longitudinal gene expression. Sri
Lankan Journal of Applied Statistics, 5(4), 117-126.
Details
Package: TSPC
Type: Package
Version: 2.0
Date: 2014-11-25
License: GPL
LazyLoad: yes
Author(s)
Yuping Zhang <yupingz@stanford.edu>
References
Zhang Y, Tibshirani RJ, Davis RW. Predicting patient survival from longitudinal gene expression.
Stat Appl Genet Mol Biol. 2010;9(1):Article41. Epub 2010 Nov 22.
tspc.cv Cross-validation
Description
This function uses a form of cross-validation to estimate the optimal feature threshold in supervised
principal components
Usage
tspc.cv(fit, data, seed = 123, topfea = TRUE, n.topfea = 1000, n.threshold = 20, n.fold = NULL, folds = NULL, n.components = 1, min.features = 5, max.features = nrow(data$x[[1]]), type = "survival")
tspc.cv 3
Arguments
fit Object returned by tspc.train
data Data object of form described in tspc.train documentation
seed A Numeric number
topfea If it is TRUE, then the tuning paparmeter is the number of features
n.topfea Maximum number of features used as the tuning parameter
n.threshold Number of thresholds, when using the number of thresholds as a tuning param-
eter
n.fold Number of cross-validation folds
folds Lists of indices of cross-validation folds (optional)
n.components Number of cross-validation components to use: 1,2 or 3.
min.features Minimum number of features to include, in determining range for threshold.
Default 5.
max.features Maximum number of features to include, in determining range for threshold.
Default is total number of features in the dataset.
type "survival" or "regression"
Details
This function uses a form of cross-validation to estimate the optimal feature threshold.
Value
list(thresholds = thresholds, n.threshold = n.threshold, nonzero = nonzero, scor = scor, scor.lower
= scor.lower, scor.upper = scor.upper, folds = folds, n.fold = n.fold, featurescores.folds = fea-
turescores.folds, type = type)
thresholds Vector of thresholds considered
n.threshold Number of thresholds
nonzero Number of features exceeding each value of the threshold
scor Full CV scores
scor.lower Full CV scores minus one standard error of scores
scor.upper Full CV scores plus one standard error of scores
folds Indices of CV folds used
n.fold Number of folds used in the cross-validation
featurescores.folds
Feature scores for each fold
type problem type
Author(s)
Yuping Zhang
4tspc.plotcv
Examples
x = list()
for(i in 1:2){
set.seed(i+123)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
x = list()
for(i in 1:2){
set.seed(i+133)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data.test = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
fit = tspc.train(data, data.test, type="survival")
cv.obj = tspc.cv(fit$fit.obj, data, type="survival", n.fold=2)
tspc.plotcv Plot output from tspc.cv
Description
Plots pre-validation results from plotcv, to aid in choosing best threshold
Usage
tspc.plotcv(object)
Arguments
object Object returned by tspc.cv
Author(s)
Yuping Zhang
Examples
x = list()
for(i in 1:2){
set.seed(i+123)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
tspc.predict 5
data = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
x = list()
for(i in 1:2){
set.seed(i+133)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data.test = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
fit = tspc.train(data, data.test, type="survival")
cv.obj = tspc.cv(fit$fit.obj, data, type="survival", n.fold=2)
tspc.plotcv(cv.obj)
tspc.predict Form principal components predictor from a trained tspc object
Description
Computes supervised principal components, using scores from "object"
Usage
tspc.predict(object, data, newdata, threshold, n.components = 3, prediction.type = c("continuous", "discrete", "nonzero"), n.class = 2)
Arguments
object Object fit.obj returned by tspc.train
data List of projection of training data returned by tspc.train, object proj.obj$wdata.train
newdata List of projection of test data returned by tspc.train, object proj.obj$wdata.test
threshold Threshold for scores.
n.components Number of principal components to compute. Should be 1,2 or 3.
prediction.type
"continuous" for raw principal component(s); "discrete" for principal compo-
nent categorized in equal bins; "nonzero" for indices of features that pass the
threshold
n.class Number of classes into which predictor is binned (for prediction.type="discrete"
Value
list(v.pred = out, u = x.sml.svd$u, d = x.sml.svd$d, which.features = which.features, v.pred.1df
= v.pred.1df, n.components = n.pc, coef = result$coef, call = this.call, prediction.type = predic-
tion.type)
v.pred Supervised principal componients predictor
uU matrix from svd of weighted feature matrix
dsingual values from svd of weighted feature matrix
which.features Indices of features exceeding threshold
n.components Number of supervised principal components requested
6tspc.project
Author(s)
Yuping Zhang
Examples
x = list()
for(i in 1:2){
set.seed(i+123)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
x = list()
for(i in 1:2){
set.seed(i+133)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data.test = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
fit = tspc.train(data, data.test, type="survival")
predict.obj<- tspc.predict(fit$fit.obj, fit$proj.obj$data.train, fit$proj.obj$data.test, threshold=1.0, n.components=1)
tspc.project Project time-course gene expression to weighted gene expression
Description
Project time-course gene expression to weighted gene expression
Usage
tspc.project(data, data.test, type = c("survival", "regression"))
Arguments
data List of training data, of form described in tspc.train documentation
data.test List of test data, of form described in tspc.test documentation
type Problem type: "survival" for censored survival outcome, or "regression" for sim-
ple quantitative outcome.
tspc.train 7
Value
list(data.train = wdata.train, data.test = wdata.test)
data.train Projection of training data
data.test Projection of test data
Author(s)
Yuping Zhang
tspc.train Prediction using time-course gene expression
Description
Does prediction of a quantitative regression or survival outcome, using the time-course gene ex-
pression.
Usage
tspc.train(data, data.test, type = c("survival", "regression"), s0.perc = 0.5)
Arguments
data Data object with components x- a list of p by n matrix of features, one obser-
vation per column, one matrix per time point; y- n-vector of outcome measure-
ments; censoring.status- n-vector of censoring censoring.status (1= died or event
occurred, 0=survived, or event was censored), needed for a censored survival
outcome.
data.test Data object with components x- a list of p by n matrix of features, one obser-
vation per column, one matrix per time point; y- n-vector of outcome measure-
ments; censoring.status- n-vector of censoring censoring.status (1= died or event
occurred, 0=survived, or event was censored), needed for a censored survival
outcome.
type Problem type: "survival" for censored survival outcome, or "regression" for sim-
ple quantitative outcome.
s0.perc Factor for denominator of score statistic, between 0 and 1: the percentile of
standard deviation values added to the denominator. Default is 0.5 (the median)
Value
proj.obj projection of training data and test data
fit.obj fitted object using training data
Author(s)
Yuping Zhang
8tspc.train
References
Zhang Y, Tibshirani RJ, Davis RW. Predicting patient survival from longitudinal gene expression.
Stat Appl Genet Mol Biol. 2010;9(1):Article41. Epub 2010 Nov 22.
Examples
x = list()
for(i in 1:2){
set.seed(i+123)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
x = list()
for(i in 1:2){
set.seed(i+133)
x[[i]] = matrix(rnorm(500*100), ncol=100)
}
y = sample(c(5:100), size=100, replace=TRUE)
censoring = sample(c(0,1), size=100, replace=TRUE)
data.test = list(x = x, y=y, censoring.status=censoring, genenames = as.character(paste("gene", c(1:500), sep="")), geneid = as.character(paste("gene", c(1:500), sep="")))
obj = tspc.train(data, data.test, type="survival")
