CNVassoc (version 1.4)

getQualityScore: Computes a quality score for a CNV fit

Description

This function provides different types of measurements of uncertainty after CNV calling

Usage

getQualityScore(x, ...)
## S3 method for class 'default':
getQualityScore(x, sds, w, type, iter = 10000, threshold = 0.1, ...)
## S3 method for class 'cnv':
getQualityScore(x, type = "class", iter = 10000, threshold = 0.1, ...)

Arguments

x

and object of class cnv or means vector of intensity signal for each copy number status

...

further arguments passed to or from getQualityScore methods

type

the type of quality score measurement computed. Possible values are "class", "CNVtools" or "CANARY" (see Details)

iter

number of iterations when type=''class'' or type=''CANARY'' is specified

threshold

a value to compute the proportion of sample individuals with confidence score bigger than it (see Details)

sds

standard deviations vector of intensity signal for each copy number status

w

copy number status proportions vector

Value

• An object of class getQualityScore with a single number of quality score.

Details

The quality scores measures how well the clusters are separated. If compares the locations of the means with the standard error for each pair of adjacent cluster. Obviously, except for probability of good classification (type="class"), the lower quality score the highest uncertainty. There are 3 possible types of quality score measurements: "class": probability of good classification), "CNVtools": the score defined in 'CNVtools' package) and "CANARY": proportions of sample individuals with confidence score bigger than threshold. The confidence score is defined as the ratio between the second biggest copy number call probability divided by the biggest one.

Examples

data(dataMLPA)
CNV<-cnv(x = dataMLPA$Gene2, threshold.0 = 0.01, mix.method = "mixdist")
getQualityScore(CNV,type="class")
getQualityScore(CNV,type="CNVtools")
getQualityScore(CNV,type="CANARY")