Premium
Queries of MALDI‐imaging global datasets identifying ion mass signatures associated with tissue compartments
Author(s) -
Fehniger Thomas E.,
Suits Frank,
Végvári Ákos,
Horvatovich Peter,
Foster Martyn,
MarkoVarga György
Publication year - 2014
Publication title -
proteomics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.26
H-Index - 167
eISSN - 1615-9861
pISSN - 1615-9853
DOI - 10.1002/pmic.201300431
Subject(s) - mass spectrometry imaging , computer science , maldi imaging , volume (thermodynamics) , metadata , sampling (signal processing) , visualization , ion , pattern recognition (psychology) , position (finance) , data mining , computational biology , mass spectrometry , artificial intelligence , chemistry , biology , physics , computer vision , matrix assisted laser desorption/ionization , chromatography , organic chemistry , filter (signal processing) , adsorption , quantum mechanics , desorption , finance , economics , operating system
Scanning MS by MALDI MS imaging (MALDI‐MSI) creates large volumetric global datasets that describe the location and identity of ions registered at each sampling location. While thousands of ion peaks are recorded in a typical whole‐tissue analysis, only a fraction of these measured molecules are purposefully scrutinized within a given experimental design. To address this need, we recently reported new methods to query the full volume of MALDI‐MSI data that correlate all ion masses to one another. As an example of this utility, we demonstrate that specific ion peak m/z signatures can be used to localize similar histological structures within tissue samples. In this study, we use the example of ion peak masses that are associated with tissue spaces occupied by airway bronchioles in rat lung samples. The volume of raw data was preprocessed into structures of 0.1 mass unit bins containing metadata collected at each sampling position. Interactive visualization in ParaView identified ion peaks that especially showed strong association with airway bronchioles but not vascular or parenchymal tissue compartments. Further iterative statistical correlation queries provided ranked indices of all m/z values in the global dataset regarding coincident distributions at any given X , Y position in the histological spaces occupied by bronchioles The study further provides methods for extracting important information contained in global datasets that previously was unseen or inaccessible.