Computational annotation of plant metabolomics profiles via a novel network-assisted approach

Mass spectrometry (MS) has become the analytical method of choice in plant metabolomics. Nevertheless, metabolite annotation remains a major challenge and implies the integration of structural searches in compound libraries with biological knowledge inferred from metabolite regulation studies. Here...

Full description

Saved in:
Bibliographic Details
Published in:Metabolomics 2013-08, Vol.9 (4), p.904-918
Main Authors: Gaquerel, Emmanuel, Kuhl, Carsten, Neumann, Steffen
Format: Article
Language:English
Subjects:
Biochemistry
Biomedical and Life Sciences
Biomedicine
Cell Biology
Developmental Biology
Life Sciences
Molecular Medicine
Nicotiana attenuata
Original Article
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Mass spectrometry (MS) has become the analytical method of choice in plant metabolomics. Nevertheless, metabolite annotation remains a major challenge and implies the integration of structural searches in compound libraries with biological knowledge inferred from metabolite regulation studies. Here we propose a novel integrative approach to process and exploit the rich structural information contained in in-source fragmentation patterns of high-resolution LC–MS profiles. In this approach, a correlation matrix is first calculated from individual mass features extracted by xcms processing. Mass feature co-regulation patterns corresponding to metabolite in-source fragmentation are then detected and assembled into compound spectra using the R package CAMERA and processed for in silico fragment-based structure elucidation using MetFrag. We validate the performance of this approach for the rapid annotation of the twelve largest compound spectra, including four O- acyl sugars and six 17-hydroxygeranyllinalool diterpene glycosides in metabolic profiles of insect-attacked Nicotiana attenuata leaves. Additionally, we demonstrate the power of refining MetFrag metabolite annotations based on co-regulation patterns between known and unknown compounds in the correlation matrix and proposed structural annotations on two previously un-characterized O- acyl sugars. In summary, this novel approach facilitates compound annotation from in-source fragmentation patterns using correlation between intensities of mass features of one or several metabolites. Additionally, this analysis provides further support that insect herbivory activates major metabolic reconfigurations in N. attenuata leaves.
ISSN:1573-3882
1573-3890
DOI:10.1007/s11306-013-0504-2