methods: Methods for patient-specific outlier detection

In dbottomly/pod: Tools for patient-specific expression outlier detection

Description

Implementations of the outlying degree and Zscore methodologies desribed in Bottomly et al 2013.

Usage

## S4 method for signature 'ExpressionSet'
outlyingDegree(obj, k, type=c("non.weight", "weight.before", "weight.after"), weight.func=default.weight.func)
## S4 method for signature 'matrix'
outlyingDegree(obj, k, type=c("non.weight", "weight.before", "weight.after"), weight.func=default.weight.func) 
## S4 method for signature 'ExpressionSet'
zScore(obj, robust=FALSE) 
## S4 method for signature 'matrix'
zScore(obj, robust=FALSE)

Arguments

obj	Either a matrix with rows corresponding to genes and columns corresponding to samples or an ExpressionSet object
k	An integer value that can range from 1 to ncol(obj)-1
type	One of either "non.weight", "weight.before" or "weight.after" which corresponds to the standard outlying degree method, whether to apply the weights specified in weight.func before computing the distance or after computing the distance respectively.
weight.func	A function returning a square matrix containing the weights corresponding to the sample-sample dissimilarities.
robust	Whether a robust Zscore should be computed e.g. (x-median(x))/mad(x)

Author(s)

Daniel Bottomly

Examples

data(simEset)
out.deg.eset <- outlyingDegree(simEset, k=1)
cur.mat <- exprs(simEset)
out.deg.mat <- outlyingDegree(cur.mat, k=1)
all.equal(out.deg.eset, out.deg.mat)

zscore.eset <- zScore(simEset)
zscore.mat <- zScore(cur.mat)

all.equal(zscore.eset, zscore.mat)

dbottomly/pod documentation built on May 15, 2019, 1:23 a.m.