In michaelwitting/masstrixR: Annotation of MS based metabolomics data

knitr::opts_chunk$set(echo = TRUE)

Introduction

Analysis of MS^2^ data can involve several steps beyond matching with a database. Different approaches can be used to analyze tandem MS spectra. One of the most common is to search for known neutral losses from the precursor masses, specific fragment m/z or mass differences between fragments. masstrixR offers functions to execute all of this steps. The example data contains spectra of ceramides from Hänel et al. (2019, under review) detected in C. elegans. These ceramides contain a unusual C17iso sphingoid base.

First we read the spectra from the .mgf file using functions from the MSnbase package. masstrixR uses Spectrum2 for working with MS^2^ spectra. Individual spectra are best stored in a Spectra object which allows to add arbitary annotations using the mcols() function.

# load required library
library(masstrixR)
library(MSnbase)

# get example file from package
mgfFile <- system.file("extdata", "exampleData\\Celegans_lipids_ms2\\ceramides.mgf", package = 'masstrixR')

# read mgf file with MS2 spectra and isolate all MS2 spectra a list
ms2data <- readMgfData(mgfFile)
ms2spectra <- Spectra(spectra(filterMsLevel(ms2data, msLevel = 2)))

# working with logical statements for fragments
mcols(ms2spectra)$waterLoss <- unlist(lapply(ms2spectra, containsNeutralLossIon, neutralLossMass = 18.010565))
mcols(ms2spectra)$c17sphingoid <- unlist(lapply(ms2spectra, containsProductIon, productIonMz = c(250.2529, 268.2635, 238.2530), multiplePi = "all"))
mcols(ms2spectra)$c27ohFa <- unlist(lapply(ms2spectra, containsFragmentDifference, fragmentMassDifference = 392.401816))

# check the invidual columns
mcols(ms2spectra)$c17sphingoid
mcols(ms2spectra)$c27ohFa

# make combined searches
selectedSpectrum <- ms2spectra[which(mcols(ms2spectra)$waterLoss & 
                   mcols(ms2spectra)$c17sphingoid &
                   mcols(ms2spectra)$c27ohFa)]

# plot
plotSpectrum(selectedSpectrum[[1]], plotIt = TRUE, highlight = TRUE, highlightMz = c(250.2529, 268.2635, 238.2530))

# get example file from package
ms2db <- system.file("extdata", "exampleData\\Celegans_lipids_ms2\\wormLipidDb_MS2.sqlite", package = 'masstrixR')

# adducts to search for
adducts <- c("[M+H]+", "[M+Na]+")

# empty data frame for results
result <- data.frame()

for(i in seq_along(ms2spectra)) {

    # create empty Spectra object for results
  searchResult <- new("Spectra")

  # perform precursor search
  # iterate over adducts
  for(adduct in adducts) {
    searchResultClipboard <-
      masstrixR::searchByPrecursor(
        precursorMz(ms2spectra[[i]]),
        ms2db,
        precursorType = adduct,
        mzTol = 0.005
      )

    # if 1 or more results were found, add to searchResult
    if(length(searchResultClipboard) > 0) {
      searchResult <- append(searchResult, searchResultClipboard)
    }
  }

  result <- rbind.data.frame(result, createResultsSet(ms2spectra[[i]], searchResult, prefix = "test", storePlot = FALSE))
}