Re: [Metware-devel] new xcmsFragments contains filterLine +acquisitionNumbers +last mz filterLine

[Malick H. <heu...@ch...>]

An important set of commands that can be used to see which
Peaks are ommited

# Assuming that you have created an object xraw with the xcmsRaw  
function call. #This command filters out all last mz's from your  
xraw@filterLine slot

> bound <- unique(as.numeric( sub("([^@]*).*","\\1", (grep("@",  
> unlist(strsplit(levels(xraw@filterLine), " ")), value=TRUE)))))

# Assuming that you have created an object xfrag with the  
xcmsFragments #function call. This command compares your column  
xfrag@peaks[,5] (storing the #xf@peaks mz values ) with all the  
filterLine mz's and keeps the mz in noise #that are not present in  
xfrag@peaks[,5]. The second command converts the noise #back to  
numeric values


> noise   <-  
> setdiff(as.character(bound),as.character(round(xfrag@peaks[,5],2)));
> noise <- as.numeric(noise)

# Now you need to know a priori howmany peaks are returned by xf@peaks  
and you #can cbind the xfrag@peaks[,5] column to the xfrag@peaks[,6]  
column while #omitting the ms1 peaks. So that they properly bind to  
each other. I have for #example 48 peaks in xf@peaks and 5 ms1 peaks,  
so I start the cbind for #xf@peaks[,6] as follow:

> cbind(round( (xfrag@peaks)[xfrag@peaks[,2],5],2) , xfrag@peaks[5:48,6])


#this allows to see if the filterLine mz and xf#peaks mz are the same
#noise be called to see which peaks are skipped ;-)



Quoting Malick Heuvel <heu...@ch...>:

> With this message I ask some attention for a new
> xcmsFragments I've added 5 new columns in the xcmsFragments
> xf@peaks table the:
> * CID, * filterLine, * AquisitionNumber of a FilterLine, * last mz of
> a filterLine
>
> In this way we can compare for a Peak if its corresponding
> msnParentPeak's mz value correspond to last mz in FilterLine. Still a
> note to be aware of.
>
> There are cases that an mzxml file has more last mz's in
> unique(FilterLines) which are not present in xf@peaks. How to deal
> with those missing peaks are they noise peaks when to state that they
> are noise.
>
> The aquisitionNumber in xf@peaks tells for each peak from which scan
> it was derived and corresponds with the  "Scan num=" in your mzxml file.
> The same Scan in mzxml has an filterLine in xf@peaks, which should
> correspond with the mz in column "last mz of a filterLine" in
> xf@peaks. I verified that this was the case for an FTMS experiment
> with unusual settings. Please your attention and comments. I want to
> be sure that its failsafe
>
>
> Regards
>
> Malick A.D Heuvel
> Netherlands Metabolomics Centre
> Gorlaeus Laboratories
> PO Box 9502
> 2300 RA Leiden
> The Netherlands



Malick A.D Heuvel
Netherlands Metabolomics Centre
Gorlaeus Laboratories
PO Box 9502
2300 RA Leiden
The Netherlands