readIllumina: Read bead-level Illumina data

In markdunning/beadarray: Quality assessment and low-level analysis for Illumina BeadArray data

Description

Reads bead-level ouput by Illumina's BeadScan software.

Usage

  readIllumina(dir = ".", useImages = FALSE, illuminaAnnotation = NULL, sectionNames = NULL, metricsFile = "Metrics.txt", forceIScan = FALSE, dec = ".", sampleSheet="sampleSheet.csv",...)

Arguments

dir	Directory from which to read the data, the default being the current working directory. If this is set to NULL, a sample sheet can be used (see below)
useImages	Boolean value specifying if bead intensities should be read directly from the source text files or calculated using the bead centre coordinates and image files. More details on the intensity calculation used are given below.
illuminaAnnotation	Character string specifiying the Illumina platform on which the data were generated. This is optional and will help to automate some analyses on expression data. Currently the choices for this argument are Humanv4, Humanv3, Humanv2, Humanv1, Mousev2, Mousev1, Mousev1p1, Ratv1.
sectionNames	Optional vector of character strings corresponding to section names to be read in. Typically these are the 10 digit numeric ID of Illumina chip followed by an underscore and capital letter
metricsFile	Optional name of a metrics file to be read in from the directory. Defaults to Metrics.txt. See details below for more information.
forceIScan	Only necessary for data from iScan machines. When TRUE, this argument forces beadarray to read data from files with the "perBeadFile.txt" extension, rather than seeking a file per swath. Use this if you are receiving the error "iScan data detected. This must first be processed with the function processSwathData()", but don't have access to .locs and .tif files for the array.
dec	Decimal point character. This should be a character string containing just one single-byte character.
sampleSheet	A comma-separated file that can be used to define the sections that make up the experiment to be analysed. See readSampleSheet for the specifications of this file
...	Other arguments to pass when reading bead-level text files.

Details

The bead-level data can be generated by any Illumina assay (expression, genotyping, methylation) and via BeadChips or Sentrix Array Matrix. However, some operations within the package are optimised for expression data. For optimal performance, BeadScan needs to be modified to output coordinates for each bead and to include outliers. See http://www.compbio.group.cam.ac.uk/Resources/illumina/index.html for details.

If present, the function will automatically read the following files from the directory

.txt Text file that lists the ID, coordinates and intensity for every decoded bead on an array section. The intensities have been subjected to a local background correction. If useImages = FALSE these intensities will be used as a starting point for analysis

.sdf Illumina\'s Sample Description File for the entire chip or SAM. This is used within beadarray to determine the physical properties of a section.

.locs Locations of all beads on the array (i.e. including all those that could not be decoded)

Metrics.txt Illumina\'s metrics that are produced at the time of scanning. Identification of the Metrics file is defined by the metricsFile argument. Internally, this argument is passed to grep. The default will thus find other items such as "Metrics.txt2" or "OtherMetrics.txt". To avoid these being read you can use a more specific regex such as "^Metrics.txt$".

Separate functionality exists to read and manipulate TIFF images that may be found in the same directory. See readTIFF.

If useImage is set to TRUE, intensities are extracted from the TIFF images using the bead-centre coordinates provided in the text file. Foreground values are calculated from a sharpened version of the image as described in Kuhn et al. Background values are calculated as the median of the 5 lowest pixels in a 17x17 pixel square around the bead-centre, detailed in Smith et al.

Value

Returns an object of class beadLevelData-class

Author(s)

Mike Smith, Mark Dunning, Andy Lynch

References

Kuhn K, Baker SC, Chudin E, Lieu MH, Oeser S, Bennett H, Rigault P, Barker D, McDaniel TK, Chee MS. A novel, high-performance random array platform for quantitative gene expression profiling. Genome Research. 2004;14(11):2347-2356

Smith ML, Dunning MJ, Tavare S, Lynch AG. Identification and correction of previously unreported spatial phenomena using raw Illumina BeadArray data. BMC Bioinformatics. 2010;11:208

markdunning/beadarray documentation built on May 9, 2019, 8:35 a.m.