gcrma.adjust.nomm: GCRMA pre-processing of arrays with custom CDF format
In jotsetung/generalgcrma: GCRMA background adjustment for Affymetrix Tiling, Exon and Gene ST microarrays.

Description Usage Arguments Details Value Author(s) References See Also

gcrma.adjust.nomm is a wrapper function for functions in the gcrma package that allows to GCRMA-pre-process Affymetrix arrays with non-typical CDF package and layout (e.g. Affymetrix Exon arrays with custom CDF files). The function performs all steps of the GCRMA preprocessing: optical background correction, NSB background adjustment, quantile normalisation and RMA summarisation.

gcrma.adjust.nomm(x, optical.correct = TRUE, verbose = TRUE,
affinity.spline.coefs, NCprobe, k = 0.5, GSB.adjust = TRUE, NC.pattern =
"bg", normalize.quantiles = TRUE, normalize.method="quantiles",
normalize.log2=FALSE, summarize = FALSE, summarize.method="rma", cl,
mc.cores, ret.residual.aggr=FALSE, residual.aggr.fun=function( z ){
mean( abs( z ) ) } )

`x`	An AffyBatch object
`optical.correct`	If the data should be optical background adjusted.
`verbose`	Print progress information.
`affinity.spline.coefs`	The affinity spline coefficients for the GCRMA NSB adjustment. The package provides pre-calculated coefficients for human Exon arrays, that can be loaded with data( "affinity.spline.coefs.exon" ). Alternatively they can be calculated from the data using the `base.profiles.nc` or `base.profiles` from the `gcrma` package.
`NCprobe`	Indices of non-specific-background control probes within the PM data (not indices of probes in the matrix returned by the functions `intensity` or `exprs`, but indices of the probes in the intensity matrix returned by the `pm` function). The function tries to find NSB control probes within the data using a pattern search on the probe set ids defined in the CDF environment.
`k`	A tuning factor
`GSB.adjust`	If probe effects in specific binding should be adjusted.
`NC.pattern`	Pattern that is used to search for NCprobes in the CDF environment's probe set names.
`normalize.quantiles`	If the data should be quantiles normalised.
`normalize.method`	Which normalization method should be used. At the moment only `quantiles` normalization is implemented.
`normalize.log2`	Wheter the normalization should be performed in log2 or natural scale.
`summarize`	If the probe intensities should be summarised to probe set intensities (using the `rma` function/model). If FALSE an AffyBatch will be returned by the function with the background adjusted probe intensities.
`summarize.method`	The summarization method/function to be used. Either rma or plm for standard RMA summarization using the median polish algorithm or summarization with the fitPLM function from the affyPLM package.
`cl`	Optional paramter. A cluster object created by the `makeCluster` function from the `snow` package. If submitted a parallelized version of this function is used.
`mc.cores`	Optional paramter. Number of cores (CPUs) on which the background adjustment should be performed. Provides, similar to parameter `cl`, a significant speed-up of the calculation. Requires that the `multicore/parallel` package is installed.
`ret.residual.aggr`	Only for summarization using the fitPLM function (see parameter `summarize.method` above). Whether probe residual values should be aggregated per probe set. Note that if set to `TRUE` a `ResExpressionSet` will be returned with the aggregated residuals available in slot `residuals`.
`residual.aggr.fun`	A function that can be used to aggregate (combine/summarize) the probe residuals of a probe set.

The function estimates the affinity spline coefficients using the expression intensities and sequences of the NSB background probes if affinity.spline.coefs is not submitted.

An ExpressionSet with GCRMA preprocessed probe set intensities (if summarize=TRUE) or an AffyBatch object with GCRMA background adjusted probe intensity values (in natural scale). If ret.residual.aggr is set to TRUE a ResExpressionSet object is returned, that is essentially identical to an ExpressionSet, with an additional slot with a matrix (same dimensions than the expression matrix) of the aggregated probe residuals per probe set.

Johannes Rainer

Rainer J, Lelong J, Bindreither D, Mantinger C, Ploner C, Geley S, Kofler R. (2012) Research resource: transcriptional response to glucocorticoids in childhood acute lymphoblastic leukemia. Mol Endocrinol. 2012 Jan;26(1):178-93.

http://bioinfo.i-med.ac.at

gcrma.bkg.correct.matrix, gcrma.compute.affinities.local.pmonly

jotsetung/generalgcrma documentation built on May 19, 2019, 9:41 p.m.