fixRI: Fixing Retention Time Index Correction
In TargetSearch: A package for the analysis of GC-MS metabolite profiling data

Description Usage Arguments Details Value Note Author(s) See Also Examples

This function can be used to correct the detected retention time index (RI) markers or to manually force their location to specific retention times if, for example, the RI markers were not co-injected with the biological samples.

1	fixRI(samples, rimLimits, RImatrix=NULL, sampleNames=NULL, quiet=TRUE)

`samples`	A `tsSample` object created by `ImportSamples` function.
`rimLimits`	A `tsRim` object. See `ImportFameSettings`.
`RImatrix`	Optional. A retention time matrix of the found retention time markers that was obtained after running `RIcorrect`.
`sampleNames`	Optional. A character vector naming the samples that are to be RI corrected.
`quiet`	Logical. Do not print a list of converted files.

Sometimes the retention index limits (see ImportFameSettings) are not set correctly and you will have to run the peak detection and RI correction function (RIcorrect) again, which may take a long time specially if there are many samples.

Instead, a simple approach is to fix the RI limits and use this function to correct the generated RI files. Since these files are much smaller than CDF files (chromatograms), this runs much faster.

Other possibility is that the time positions of one or more RI markers were wrongly detected because there was just simply no peak or the RI markers where not co-injected in some samples. You could manually force the locations of the RI markers. This case is discussed in the “RICorrection” vignette.

The behavior of this function depends on whether the parameter RImatrix is NULL or not. If NULL, the RI markers will be searched again (using the settings of rimLimits parameters, which you should have already fixed) and the resulting values will be used to correct the RI files. If it is a numeric matrix, then these values will be used to correct the RI files. Note that this matrix dimensions are exactly m samples (rows) times n (columns) RI markers.

sampleNames controls which samples will be corrected. If NULL then all samples will be corrected. It could be character vector (sample names) or a numeric vector representing the sample indexes.

Invisible NULL. It prints the corrected samples if quiet is FALSE.

In case the RI files are in text format and their column names are not standard (for example, when the files were generated with another software), use the global option 'TS_RI_columns' or transform the RI files to TargetSearch binary format. See the documentation in function text2bin.

Alvaro Cuadros-Inostroza

RIcorrect, FAMEoutliers,ImportSamples, ImportFameSettings

require(TargetSearchData)
# import refLibrary, rimLimits and sampleDescription.
data(TSExample)
# get the CDF files
cdfpath <- file.path(find.package("TargetSearchData"), "gc-ms-data")

# select a subset of samples
smp <- sampleDescription[1:4]

# update the CDF path
CDFpath(smp) <- cdfpath

# make a copy of the RI markers object
rim <- rimLimits

# mess up the limits of marker 3 (real value is 369 seconds app.)
rimLimits(rim)[3,] <- c(375, 400)

# run RIcorrect (skip CDF-4 conversion)
RImat <- RIcorrect(smp, rim, writeCDF4path=FALSE,
           Window = 15, IntThreshold = 50)

# fix the limits of marker 3
rimLimits(rim)[3,] <- c(360, 400)

# you could run again RIcorrect, but this is faster
fixRI(smp, rim)

# get the RI matrix
RImat <- riMatrix(smp, rim)

# compare the values with the real ones (previously stored in RImatrix)
stopifnot( all.equal(RImat, RImatrix[,1:4], tolerance=1e-8) )

# manual adjustment or RI markers for sample 3.
# Warning: this is just an example to illustrate how to use this function.
#          don't do this unless you know what you're doing.
RImat[,3] <- c(252, 311, 369)
fixRI(smp, rim, RImat, 3)