allocatePrecursor2mz: allocatePrecursor2mz: Join two data sources

In PlantDefenseMetabolism/MetabolomicTools: Navigating mass spectral similarity in high-resolution MS/MS metabolomics data

Description

Allocates precursor ions to candidate m / z values based on minimal distance of m / z and deviance of rt based on an objective function

Usage

allocatePrecursor2mz(sd01, sd02, kNN = 10, mzCheck = 1, rtCheck = 30, 
     mzVsRTbalance = 10000, splitPattern = "_", splitInd = 2)

Arguments

sd01	is the output of the XCMS and CAMERA processing and statistical analysis and XCMS and CAMERA scripts (see Li et al. 2015 and vignette for further information)
sd02	data.frame with idMS/MS deconvoluted spectra with fragment ions (m/z, retention time, relative intensity in %) and the corresponding peak correlation group of the precursor ion. sd02 has to have at least four columns: a column 'mz', 'rt', 'intensity' and 'id'
kNN	numerical, number of k-nearest neighbours based on deviation from m/z (i.e. the k entries with the smallest deviation)
mzCheck	numerical, maximum tolerated distance for m/z (strong criterion here)
rtCheck	numerical, maximum tolerated distance for retention time
mzVsRTbalance	numerical, multiplicator for mz value before calculating the (euclidean) distance between two peaks, high value means that there is a strong weight on the deviation m/z value
splitPattern	character, character vector to use for splitting, see ?strsplit for further information
splitInd	numeric, extract precursor mz at position splitInd

Details

This function combines different data sources. convertExampleDF is a data.frame which comprises information on a specific metabolite per row stating the average retention time, average m/z, the name of the metabolite, the adduct ion name, the spectrum reference file name and additional information (here: TRIO/LVS). allocatePrecursor2mz uses data.frames of the kind of sd01\_outputXCMS and sd02\_deconvoluted to create a data.frame of the kind of convertExampleDF. Allocation of precursor ions to candidate m/z values is based on minimal distance of m/z and deviance of retention time based on an objective function. We can specify threshold values for m/z and retention time to be used in allocatePrecursor2mz, as well as the number of neighbours based on deviation from m/z values. Also, we can specify the weight to base the selection on the m/z compared to the retention time (mzVsRTbalance). This might be useful because m/z values might differ less than the retention time in sd01\_outputXCMS and sd02\_deconvoluted. Please note, that it might be problematic to compare sd01\_outputXCMS and sd02\_deconvoluted and allocate precursor ions therefrom, especially when data were acquired under different conditions.

Value

allocatePrecursor2mz returns a data.frame containing average retention time, average mz, metabolite name, adduct ion name, spectrum reference

Author(s)

Thomas Naake, [e-mail]

References

Li et al. (2015): Navigating natural variation in herbivory-induced secondary metabolism in coyote tobacco populations using MS/MS structural analysis. PNAS, 112, E4147–E4155, 10.1073/pnas.1503106112.

Examples

data("sd01_outputXCMS", package = "MetCirc")
data("sd02_deconvoluted", package = "MetCirc") 
data("convertExampleDF", package = "MetCirc")
allocatePrecursor2mz(sd01 = sd01_outputXCMS, sd02 = sd02_deconvoluted, 
     kNN = 10, mzCheck = 1, rtCheck = 30, mzVsRTbalance = 10000, splitPattern = " _ ", splitInd = 2)

PlantDefenseMetabolism/MetabolomicTools documentation built on May 8, 2019, 2:53 p.m.