groupCorr {CAMERA} | R Documentation |
Peak grouping after correlation information into pseudospectrum groups for an xsAnnotate object. Return an xsAnnotate object with grouping information.
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 |
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 |
... |
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