1 Introduction

This document describes how to use xcms for processing of Tandem-MS and MS\(^{n}\) data from e.g. triple-quad, QTOF, ion trap or orbitrap mass spectrometers. This uses still the old user interface.

library(xcms)
library(msdata)

## Disable parallel processing
register(SerialParam())

2 Raw data file preparation

The xcms package reads full-scan LC/MS data and associated Tandem-MS and MS\(^{n}\) spectra from mzData and mzXML files (CDF does not support MS\(^{n}\)).

For the purposes of demonstration, we will use a some sample files included in the msdata package. The (truncated) raw data files are contained in the data directory. To access the mzData files, we first locate the data directory in the msdata package.

mzdatapath <- system.file("iontrap", package = "msdata")
list.files(mzdatapath, recursive = TRUE)
## [1] "DhexD4_1.mzData"  "DhexD4_2.mzData"  "DhexD4_3.mzData" 
## [4] "extracted.mzData"

3 Accessing, combining and visualising MSn spectra

The MS\(^{n}\) spectra are read into xcms just like plain LC-MS files. Inclusion of the additional scans has to be explicitely enabled. The xcmsRaw summary shows the additional content:

library(xcms)
mzdatafiles <- list.files(mzdatapath, pattern = "extracted.mzData",
                          recursive = TRUE, full.names = TRUE)
xraw <- xcmsRaw(mzdatafiles[1], includeMSn=TRUE)
## Create profile matrix with method 'bin' and step 1 ... OK
xraw
## An "xcmsRaw" object with 100 mass spectra
## 
## Time range: 1190-1309.2 seconds (19.8-21.8 minutes)
## Mass range: 500-849.9333 m/z
## Intensity range: 100008-828487000 
## 
## MSn data on  22  mass(es)
##  with  44  MSn spectra
## Profile method: bin 
## Profile step: 1 m/z (351 grid points from 500 to 850 m/z)
## 
## Memory usage: 2.71 MB

There is also a new peak picker, which actually returns all parent ions as annotated in the mzData / mzXML files. A warning will be issued if this is not present, either because of a poor converter or if xraw is read from a CDF file:

peaks <- findPeaks(xraw, method="MS1")

By giving larger windows, subsets of spectra can be returned, and these can be combined to obtain so called compound spectra containing e.g. MS\(^{2}\) to MS\(^{5}\).

4 xcmsFragments

A xcmsFragments-Object contains a Table with informations about all Peaks from MS1 to MS\(^{n}\) of one Experiment, including mz-value, retentiontime, MSlevel, and parentpeak. The data can be used for visualization or (later) computing alignments between different Experiments. The object is created as empty and can be filled with collect(). This method requires a peaktable or a xcmsSet for the MS1peaks and the original xcmsRaw for the msn-Data.

xs <- xcmsSet(mzdatafiles, method = "MS1")
xfrag <- xcmsFragments(xs)
## Create profile matrix with method 'bin' and step 1 ... OK
## 394 Peaks picked, 0 MSn-Specs ignored.
xfrag
## An "xcmsFragments" object with  394  peaks in 77 Spectra
## From Level 1 to 3 Number of Samples:  1 .
## 
## Sample 1 :
##     37 Peaks in Level 1 
##     351 Peaks in Level 2 
##     6 Peaks in Level 3 
## 
## Memory usage: 0.0289 MB

plotTree prints out a simplyfied tree of the peaks in a xcmsFragments. The user can define borders for the retentionTime and for the MassRange, he can also print the tree for a single MS1-Peak. PlotTree uses the package RgraphViz and plots the tree text-based if requested or if the package cannot be found.

Below we simply print the textual representation of the fragment tree.

plotTree(xfrag, xcmsFragmentPeakID = 6, textOnly = TRUE)
## |##6_M814T1205.2F814
## |     ##75_M814T1206.52F673.2994
## |          ##85_M814T1207.13F586.9333
## |          ##86_M814T1207.13F588.1333
## |     ##76_M814T1206.52F672.0354
## |     ##77_M814T1206.52F573.2417
## |     ##78_M814T1206.52F658.2642
## |     ##79_M814T1206.52F742.0226
## |     ##80_M814T1206.52F742.4218
## |     ##81_M814T1206.52F657.1997
## |     ##82_M814T1206.52F587.0795
## |     ##83_M814T1206.52F755.1951
## |     ##84_M814T1206.52F743.2867