groupCorr-methods: EIC correlation grouping of LC/ESI-MS data
In CAMERA: Collection of annotation related methods for mass spectrometry data

Description Usage Arguments Details Author(s) See Also Examples

Peak grouping after correlation information into pseudospectrum groups for an xsAnnotate object. Return an xsAnnotate object with grouping information.

1
2
3

  groupCorr(object,cor_eic_th=0.75, pval=0.05, graphMethod="hcs",
  calcIso = FALSE, calcCiS = TRUE, calcCaS = FALSE, psg_list=NULL, xraw=NULL, 
  cor_exp_th=0.75, intval="into", ...)

`object`	The `xsAnnotate` object
`cor_eic_th`	Correlation threshold for EIC correlation
`pval`	p-value threshold for testing correlation of significance
`graphMethod`	Clustering method for resulting correlation graph. See calcPC for more details.
`calcIso`	Include isotope detection informationen for graph clustering
`calcCiS`	Calculate correlation inside samples
`calcCaS`	Calculate correlation accross samples
`psg_list`	Vector of pseudospectra indices. The correlation analysis will be only done for those groups
`xraw`	Optional xcmsRaw object, which should be used for raw data extraction
`cor_exp_th`	Threshold for intensity correlations across samples
`intval`	Selection of the intensity values (such as "into") that should be used in the correlation analysis. See `getPeaklist` for all allowed values.
`...`	Additional parameter

The algorithm calculates different informations for group peaks into so called pseudospectra. This pseudospectra contains peaks, with have a high correlation between each other. So far three different kind of information are available. Correlation of intensities across samples (need more than 3 samples), EIC correlation between peaks inside a sample and additional the informationen about recognized isotope cluster can be included. After calculation of all these informations, they are combined as edge value into a graph object. A following graph clustering algorithm separate the peaks (nodes in the graph) into the pseudospectra.

Carsten Kuhl <ckuhl@ipb-halle.de>

calcCiS calcCaS calcPC xsAnnotate-class

 library(CAMERA)
 file        <- system.file('mzdata/MM14.mzdata', package = "CAMERA");
 xs          <- xcmsSet(file, method="centWave", ppm=30, peakwidth=c(5, 10));
 an          <- xsAnnotate(xs);
 an.group    <- groupFWHM(an);
 an.iso      <- findIsotopes(an.group); #optional step for using isotope information
 an.grp.corr <- groupCorr(an.iso, calcIso=TRUE);
 
 #For csv output
 # write.csv(file="peaklist_with_isotopes.csv",getPeaklist(an))

 #Multiple sample 
 library(faahKO)
 xs.grp       <- group(faahko)
 
 #With selected sample
 xsa          <- xsAnnotate(xs.grp, sample=1)
 xsa.group    <- groupFWHM(xsa)
 xsa.iso      <- findIsotopes(xsa.group) #optional step
 xsa.grp.corr <- groupCorr(xsa.iso, calcIso=TRUE)

 #With automatic selection
 xsa.auto     <- xsAnnotate(xs.grp)
 xsa.grp      <- groupFWHM(xsa.auto)
 xsa.iso      <- findIsotopes(xsa.grp) #optional step
 index        <- c(1,4) #Only group one and four will be calculate
 #We use also correlation across sample
 xsa.grp.corr <- groupCorr(xsa.iso, psg_list=index, calcIso=TRUE, calcCaS=TRUE)
 #Note: Group 1 and 4 have no subgroups