crlmmIllumina {crlmm}R Documentation

Genotype Illumina Infinium II BeadChip data with CRLMM

Description

This implementation of the CRLMM is especially designed for data from Illumina Infinium II BeadChips.

Usage


crlmmIllumina(RG, XY, stripNorm=TRUE, useTarget=TRUE,
      row.names=TRUE, col.names=TRUE,
      probs=c(1/3, 1/3, 1/3), DF=6, SNRMin=5,
      gender=NULL, seed=1, save.it=FALSE, load.it=FALSE,
      intensityFile, mixtureSampleSize=10^5,
      eps=0.1, verbose=TRUE, cdfName, sns, recallMin=10,
      recallRegMin=1000, returnParams=FALSE, badSNP=0.7)

Arguments

RG NChannelSet containing R and G bead intensities
XY NChannelSet containing X and Y bead intensities
stripNorm 'logical'. Should the data be strip-level normalized?
useTarget 'logical' (only used when stripNorm=TRUE). Should the reference HapMap intensities be used in strip-level normalization?
row.names 'logical'. Use rownames - SNP names?
col.names 'logical'. Use colnames - Sample names?
probs 'numeric' vector with priors for AA, AB and BB.
DF 'integer' with number of degrees of freedom to use with t-distribution.
SNRMin 'numeric' scalar defining the minimum SNR used to filter out samples.
gender 'integer' vector, with same length as 'filenames', defining sex. (1 - male; 2 - female)
seed 'integer' scalar for random number generator (used to sample mixtureSampleSize SNPs for mixture model.
save.it 'logical'. Save preprocessed data?
load.it 'logical'. Load preprocessed data to speed up analysis?
intensityFile 'character' with filename of preprocessed data to be saved/loaded.
mixtureSampleSize 'integer'. The number of SNP's to be used when fitting the mixture model.
eps Minimum change for mixture model.
verbose 'logical'.
cdfName 'character' defining the chip annotation (manifest) to use ('human370v1c', human550v3b', 'human650v3a', 'human1mv1c', 'human370quadv3c', 'human610quadv1b', 'human660quadv1a' 'human1mduov3b')
sns 'character' vector with sample names to be used.
recallMin 'integer'. Minimum number of samples for recalibration.
recallRegMin 'integer'. Minimum number of SNP's for regression.
returnParams 'logical'. Return recalibrated parameters.
badSNP 'numeric'. Threshold to flag as bad SNP (affects batchQC)

Details

Note: The user should specify either the RG or XY intensities, not both. Alternatively if crlmmIllumina has been run already with save.it=TRUE, the preprocessed data can be loaded from file by specifying load.it=TRUE and intensityFile (RG or XY are not needed in this case).

Value

A SnpSet object which contains

calls Genotype calls (1 - AA, 2 - AB, 3 - BB)
callProbability confidence scores 'round(-1000*log2(1-p))'
SNPQC SNP Quality Scores
batchQC Batch Quality Scores

along with center and scale parameters when returnParams=TRUE in the featureData slot.

Author(s)

Matt Ritchie

References

Carvalho B, Bengtsson H, Speed TP, Irizarry RA. Exploration, normalization, and genotype calls of high-density oligonucleotide SNP array data. Biostatistics. 2007 Apr;8(2):485-99. Epub 2006 Dec 22. PMID: 17189563.

Carvalho B, Louis TA, Irizarry RA. Describing Uncertainty in Genome-wide Genotype Calling. (in prep)

Examples

## crlmmOut = crlmmIllumina(RG)

[Package crlmm version 1.2.4 Index]