incrementalRMA: Incremental RMA

Description Usage Arguments Details Value Note Examples

View source: R/incrementalRMA.R

Description

Perform incremental RMA algorithm on CEL files.

Usage

1
incrementalRMA(abatch, params = NULL, calculate_error = FALSE)

Arguments

abatch

An AffyBatch-class of cel files to apply RMA to.

params

An incremental parameter list as generated by parameterizeRMA.

calculate_error

incrementalRMA errors can be calculated vs. what canonical RMA would produce. See calculateIncrementalRMAError for a discussion of the concept and interpretation.

Details

RMA is a robust multiarray average expression (see rma). It is one of the de facto approaches for processing Affymetrix GeneChip data.

One disadvantage of RMA is that parameters are calculated on the set of arrays during the normalization and probeset summarization process. This works out for studies in which all samples are available at once. In the case of adding new samples into a dataset, the entire set must be reprocessed with RMA. Making it even more challenging, this reprocessing will result in slightly different parameters. These different parameters lead in turn to slightly different gene expression measurements.

This package gains inspiration from the frma package, as well as earlier efforts including refRMA and incrementalRMA. The goal of the package is to use a set of data to pre-calculate necessary parameters for RMA. Then, when a new sample is made available these stored parameters can be applied to the new sample using the RMA approach. The result should provide gene expression estimates similar to those used when the entire dataset is used together.

Value

A ExpressionSet-class object representing the RMA normalized AffyBatch object.

Note

A test of the package is the following: store the parameters from normalizing a dataset D. Then apply incrementalRMA to each of the raw samples in D. The resulting gene expression should be identical in both cases.

Examples

1
2
3
4
## Not run: 
incrementalRMA(AffyBatch, params=list(probeEffects=(),normalizationVector=()))

## End(Not run)

steveneschrich/IncrementalRMA documentation built on Dec. 23, 2021, 5:32 a.m.