iterativeBMA (version 1.30.0)

iterativeBMAsurv-package: The Iterative Bayesian Model Averaging (BMA) algorithm for survival analysis

Description

The iterative Bayesian Model Averaging (BMA) algorithm for survival analysis is a variable selection method for applying survival analysis to microarray data..

Details

ll{ Package: iterativeBMAsurv Type: Package Version: 0.1.0 Date: 2008-3-24 License: GPL version 2 or higher } The function iterateBMAsurv.train selects relevant variables by iteratively applying the bic.surv function from the BMA package until all variables in the training data are exhausted. The variables are assumed to be pre-sorted by rank when this function is called. The function iterateBMAsurv.train.wrapper acts as a wrapper for iterateBMAsurv.train, returning the names of the selected variables and an object of class bic.surv if the iterations exhaust all variables in the training set (-1 otherwise). Again, the variables are assumed to be pre-sorted by rank, so calling this function allows users to experiment with different univariate ranking measures. The function iterateBMAsurv.train.predict.assess combines the training, prediction, and assessment phases. It returns a list consisting of the numbers of selected genes and models from the training phase, the predicted risk scores of the test samples, and the overall survival analysis statistics indicating the difference between risk groups (p-value, chi-square statistic, and variance matrix). It also writes a Kaplan-Meier survival analysis curve to file, which serves as a pictorial nonparametric estimator of the difference between risk groups. The variables are not assumed to be pre-sorted by rank when this function is called. iterateBMAsurv.train.predict.assess calls singleGeneCoxph, which ranks the genes based on their log likelihood scores using Cox Proportional Hazards Regression. iterateBMAsurv.train.predict.assess calls iterateBMAsurv.train.wrapper in its training phase, so if Cox Proportional Hazards Regression is the desired univariate ranking algorithm, then calling this function with the training and testing sets is all that is necessary for a complete survival analysis run. The function crossVal performs k runs of n-fold cross validation on a training data set, where k and n are specified by the user. crossVal calls iterateBMAsurv.train.predict.assess during each fold, so Cox Proportional Hazards Regression is the univariate ranking measure for this function.

References

Annest, A., Yeung, K.Y., Bumgarner, R.E., and Raftery, A.E. (2008). Iterative Bayesian Model Averaging for Survival Analysis. Manuscript in Progress.

Raftery, A.E. (1995). Bayesian model selection in social research (with Discussion). Sociological Methodology 1995 (Peter V. Marsden, ed.), pp. 111-196, Cambridge, Mass.: Blackwells.

Volinsky, C., Madigan, D., Raftery, A., and Kronmal, R. (1997) Bayesian Model Averaging in Proprtional Hazard Models: Assessing the Risk of a Stroke. Applied Statistics 46: 433-448.

Yeung, K.Y., Bumgarner, R.E. and Raftery, A.E. (2005) Bayesian Model Averaging: Development of an improved multi-class, gene selection and classification tool for microarray data. Bioinformatics 21: 2394-2402.

See Also

iterateBMAsurv.train, iterateBMAsurv.train.wrapper, iterateBMAsurv.train.predict.assess, singleGeneCoxph, predictiveAssessCategory, crossVal, trainData, trainSurv, trainCens, testData, testSurv, testCens

Examples

library (BMA)
library (iterativeBMAsurv)
data(trainData)
data(trainSurv)
data(trainCens)
data(testData)
data(testSurv)
data(testCens)

## Use p=10 genes and nbest=5 for fast computation
ret.bma <- iterateBMAsurv.train.predict.assess (train.dat=trainData, test.dat=testData, surv.time.train=trainSurv, surv.time.test=testSurv, cens.vec.train=trainCens, cens.vec.test=testCens, p=10, nbest=5)

## Extract the statistics from this survival analysis run
number.genes <- ret.bma$nvar
number.models <- ret.bma$nmodel
evaluate.success <- ret.bma$statistics

## Perform 1 run of 2-fold cross validation on the training set, using p=10 genes and nbest=5 for fast computation
cv <- crossVal(exset=trainData, survTime=trainSurv, censor=trainCens, diseaseType="DLBCL", noFolds=2, noRuns=1, p=10, nbest=5)