OGSA (version 1.2.0)

outCallRank: outCallRank

Description

Counts outliers by the Ghosh method and generates list objects with all outliers noted

Usage

outCallRank (dataSet, phenotype, thres= 0.05, tail='right', corr=FALSE,
offsets=NULL, names=NULL)

Arguments

dataSet

Set of matrices of molecular data

phenotype

A vector of 0s and 1s of length nSample, where 1 = case, 0 = control

thres

Alpha value

tail

A vector equal to the number of matrices with values left or right for where to find outliers

corr

Whether to correct for normal outliers

offsets

A vector equal to the number of matrices which sets the minimum value relative to normal to call outlier (corrected rank only)

names

A vector equal to the number of matrices to name molecular type of data (e.g., CNV)

Value

A list with all specific outlier calls for each molecular type in each case sample

References

Ochs, M. F., Farrar, J. E., Considine, M., Wei, Y., Meshinchi, S., & Arceci, R. J. (n.d.). Outlier Analysis and Top Scoring Pair for Integrated Data Analysis and Biomarker Discovery. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 1-1. doi:10.1109/tcbb.2013.153

D. Ghosh. (2010). Discrete Nonparametric Algorithms for Outlier Detection with Genomic Data. J. Biopharmaceutical Statistics, 20(2), 193-208.

Examples

data(ExampleData)

#set up dataSet
dataSet <- list(expr, meth,cnv)

# Set up Phenotype
phenotype <- pheno
names(phenotype) <- colnames(cnv)

# set up values for expr-meth-cnv in that order
tailLRL <- c('left', 'right', 'left')

outRankLRL <- outCallRank(dataSet, phenotype, names=c('Expr',
                             'Meth', 'CNV'), tail=tailLRL)