seabi: Calculate a performance metric based on male-female...

In schalkwyk/wateRmelon: Illumina DNA methylation array normalization and metrics

Calculate a performance metric based on male-female differences for Illumina methylation 450K arrays

Description

Calculates an area under ROC curve - based metric for Illumina 450K data using a t-test for male-female difference as the predictor for X-chromosome location of probes. The metric is 1-area so that small values indicate good performance, to match our other, standard error based metrics gcose and dmrse. Note that this requires both male and female samples of known sex and can be slow to compute due to running a t-test on every probe.

Usage

seabi(bn, stop = 1, sex, X)

Arguments

bn	a matrix of betas (default method) or an object containing betas i.e. a MethyLumiSet object (methylumi package), a MethylSet or RGChannelSet object (minfi package) or a exprmethy450 object (IMA package).
stop	partial area under curve is calculated if stop value <1 is provided
sex	a factor giving the sex of each sample (column)
X	a logical vector of length equal to the number of probes, true for features mapped to X-chromosome

Value

a value between 0 and 1. values close to zero indicate high data quality as judged by the ability to discriminate male from female X-chromosome DNA methylation.

Author(s)

[e-mail]

References

Pidsley R, Wong CCY, Volta M, Lunnon K, Mill J, Schalkwyk LC: A data-driven approach to preprocessing Illumina 450K methylation array data (submitted)

Examples

   library(methylumi)
   data(melon)
   sex  <- pData(melon)$sex
   X    <- melon@featureData@data$CHR=='X'
   seabi(betas(melon), sex=sex, X=X)

# methylumi method
   seabi(melon, sex=sex, X=X)

schalkwyk/wateRmelon documentation built on Aug. 13, 2024, 9:52 a.m.