I'm using a q-exactive plus connected to a uplc through an ESI source. After using xcms, the differport matrix contains a lot (too much) of variables with zero intensities, mainly in the samples, less so in the QC's.
Here is the script I use
xset<-xcmsSet(method="centWave", peakwidth=c(3,20), snthresh=5, mzdiff=0.01, ppm=3, prefilter=c(5,100000), noise=100000)
xset
xset2<-retcor(xset, method="obiwarp", profStep=0.1, plottype="deviation")
xset2
xset2<- group(xset2, bw=5, mzwid=0.008)
xset3<-fillPeaks(xset2)
xset3
I shall say that we previoulsy used a QTOF (Bruker microTOF Q) with of course a different xcms tuning, but never got such trouble.
Any ideas of what's going wrong with fillpeaks using he q-exactive data?