Robust Data Processing and Normalization Strategy for MALDI Mass Spectrometric Imaging

Abstract

Matrix-assisted laser desorption/ionization (MALDI) mass spectrometry imaging (MSI) provides localized information about the molecular content of a tissue sample. To derive reliable conclusions about MSI data, it is necessary to implement appropriate processing steps in order to compare peak intensities across the different pixels comprising the image. Here, we review commonly used normalization methods, and propose a rational data processing strategy, for robust evaluation and modeling of MSI data. The approach includes newly developed heuristic methods for selecting biologically relevant peaks and pixels to reduce the size of a data set and remove the influence of the applied MALDI matrix. The methods are demonstrated on a MALDI MSI data set of a sagittal section of rat brain (4750 bins, m/z = 50-1000, 111 × 185 pixels) and the proposed preferred normalization method uses the median intensity of selected peaks, which were determined to be independent of the MALDI matrix. This was found to effectively compensate for a range of known limitations associated with the MALDI process and irregularities in MS image sampling routines. This new approach is relevant for processing of all MALDI MSI data sets, and thus likely to have impact in biomarker profiling, preclinical drug distribution studies, and studies addressing underlying molecular mechanisms of tissue pathology.

Imaging mass spectrometry statistical analysis

“…Consequently the non-specific noise that is the principal contributor to a spectrum's TIC for MALDI-ToF instruments is absent. Fonville et al have investigated normalization under such conditions, using data acquired on a MALDI Q-ToF instrument [35]. They investigated the use of the total ion count of the picked peaks and also other normalization methods such as normalization to matrix-related peaks or informative peaks.…”

Chemo-informatic strategy for imaging mass spectrometry-based hyperspectral profiling of lipid signatures in colorectal cancer

“…The objectives of variance-stabilizing normalization are to remove biologically unrelated pixel-to-pixel variation in overall signal intensity and to convert multiplicative noise into additive noise for subsequent application of multivariate statistical techniques; i.e., vsn x kðixjÞ n ðixjÞ ≈ μ kðixjÞ + « kðixjÞ , [3] where vsn denotes a variance-stabilizing normalization, μ k(ixj ) is the transformed peak intensity, and « kðixjÞ is random additive noise. Here, the normalization factor has been estimated by calculating the median peak intensity, which we have shown is a more robust estimate compared with the widely cited TIC normalization method (21,23). Due to peak integration and noise-related peak filtration steps, it was assumed that the influence of background noise in the model (Eq.…”

“…This reduces mass detection accuracy and introduces biologically irrelevant spectral features, making unambiguous assignment of chemical species more difficult. In the case of normalization, the total ion current (TIC) scaling factor is frequently cited in the literature as an acceptable means of accounting for global intensity changes in a MSI dataset (20)(21)(22). However, we have recently demonstrated that the performance of this method can be compromised by single large molecular ion peak intensities (21,23).…”

“…In the case of normalization, the total ion current (TIC) scaling factor is frequently cited in the literature as an acceptable means of accounting for global intensity changes in a MSI dataset (20)(21)(22). However, we have recently demonstrated that the performance of this method can be compromised by single large molecular ion peak intensities (21,23). An additional problem inherent to MS-based analysis of complex biological mixtures is the fact that molecules present in greater intensities within a given sample will tend to exhibit larger variations when subjected to repeated measurement (23).…”

Scores for standardization of on-tissue digestion of formalin-fixed paraffin-embedded tissue in MALDI-MS imaging

“…In parallel, computational methods for MSI data pre-processing, feature extraction, (tumor) tissue classification and spatial autocorrelation have been advanced [23][24][25][26][27][28][29][30][31][32][33][34], but most of them still focus on MS imaging of non-digested human FFPE or frozen tissue [24,29,[35][36][37][38][39][40]. Classification of human tumors based on digested FFPE tissue and MALDI MSI has been described, in seminal proof-of-concept studies, for pancreatic cancer [41] and prostate cancer [42].…”