RATIONALE:The field of mass spectrometry imaging is currently devoid of standardized protocols or commercially available products designed for system suitability testing of MSI platforms. Machine learning is an approach that can quickly and effectively identify complex patterns in data and use them to make informed classifications, but there is a technical barrier to implementing these algorithms. Here we package the machine learning algorithms into a user-friendly interface to make community-wide implementation of this protocol possible. METHODS:The software package is built entirely in the Python language using the PySimpleGUI library for the construction of the interface, Pandas and Numpy libraries for data formatting and manipulation, and the Scikit-Learn library for the implementation of machine learning algorithms. Training data is collected on an instrument under clean and compromised conditions that can then be used to evaluate model performance and to train models prior to interrogating unknown samples before, during, or after experiments. RESULTS:Detailed instructions are provided for the effective use of the SLICE-MSI software package to use machine learning to evaluate instrument condition of MSI platforms. File formatting and generalizable steps are clearly described to make the implementation of this package easy for multiple labs and different MSI platform configurations. CONCLUSIONS:In this protocol, we demonstrate SLICE-MSI, a machine learning graphical user interface for efficient and easy implementation of QC instrument classification of mass spectrometry imaging platforms.

RATIONALE:While quality control (QC) and system suitability testing (SST) methods are commonly employed in mass spectrometry, the field of mass spectrometry imaging (MSI) currently lacks any universally accepted QC/SST protocols. These methods can prevent the loss of precious samples due to suboptimal instrument conditions and/or data quality, but they are more challenging to implement on MSI platforms. Herein, a panel of analytes is conveniently analyzed in a setup that reflects a typical MSI imaging experiment, and guidance is provided for downstream QC/SST evaluation. METHODS:The analyte panel will be commercially available and consists of three pairs of unlabeled (NAT) analytes and their stable isotope-labeled (SIL) analogues; a deviation from the standard procedure is also included, which incorporates a polymer to expand m/z coverage. The NAT three-plex (or four-plex with the added polymer) is analyzed as a droplet on a slide, and the SIL three-plex is doped into the electrospray solvent, isolating the NAT and SIL compounds to different source components. Datasets are collected on clean and compromised instruments to inform QC/SST software and later evaluate instrument conditions or isolated metrics of data quality. RESULTS:A procedure was created for QC/SST analysis on MSI platforms, which can be optionally paired with the freely available software Supervised Learning for Instrument Classification and Evaluation for Mass Spectrometry Imaging (SLICE-MSI) to classify the condition of the instrument. The SIL data may be monitored separately during imaging experiments for continuous evaluation of electrospray stability. The protocol highlights areas that may be adapted for other ionization sources for widespread use. CONCLUSIONS:The protocol described herein uses a panel of NAT and SIL compounds to offer an objective and accurate determination of QC/SST on MSI platforms.

RationaleMass spectrometry imaging (MSI) elevates the power of conventional mass spectrometry (MS) to multidimensional space, elucidating both chemical composition and localization. However, the field lacks any robust quality control (QC) and/or system suitability testing (SST) protocols to monitor inconsistencies during data acquisition, both of which are integral to ensure the validity of experimental results. To satisfy this demand in the community, we propose an adaptable QC/SST approach with five analyte options amendable to various ionization MSI platforms (e.g., desorption electrospray ionization, matrix-assisted laser desorption/ionization [MALDI], MALDI-2, and infrared matrix-assisted laser desorption electrospray ionization [IR-MALDESI]).MethodsA novel QC mix was sprayed across glass slides to collect QC/SST regions-of-interest (ROIs). Data were collected under optimal conditions and on a compromised instrument to construct and refine the principal component analysis (PCA) model in R. Metrics, including mass measurement accuracy and spectral accuracy, were evaluated, yielding an individual suitability score for each compound. The average of these scores is utilized to inform if troubleshooting is necessary.ResultsThe PCA-based SST model was applied to data collected when the instrument was compromised. The resultant SST scores were used to determine a statistically significant threshold, which was defined as 0.93 for IR-MALDESI-MSI analyses. This minimizes the type-I error rate, where the QC/SST would report the platform to be in working condition when cleaning is actually necessary. Further, data scored after a partial cleaning demonstrate the importance of QC and frequent full instrument cleaning.ConclusionsThis study is the starting point for addressing an important issue and will undergo future development to improve the efficiency of the protocol. Ultimately, this work is the first of its kind and proposes this approach as a proof of concept to develop and implement universal QC/SST protocols for a variety of MSI platforms.

Quality control and system suitability testing are vital protocols implemented to ensure the repeatability and reproducibility of data in mass spectrometry investigations. However, mass spectrometry imaging (MSI) analyses present added complexity since both chemical and spatial information are measured. Herein, we employ various machine learning algorithms and a novel quality control mixture to classify the working conditions of an MSI platform. Each algorithm was evaluated in terms of its performance on unseen data, validated with negative control data sets to rule out confounding variables or chance agreement, and utilized to determine the necessary sample size to achieve a high level of accurate classifications. In this work, a robust machine learning workflow was established where models could accurately classify the instrument condition as clean or compromised based on data metrics extracted from the analyzed quality control sample. This work highlights the power of machine learning to recognize complex patterns in MSI data and use those relationships to perform a system suitability test for MSI platforms.

He holds a Bachelor of Science degree in chemistry (2021) from Wofford College in Spartanburg, South Carolina.His research involves fundamental investigations to improve the IR-MALDESI-

Mass spectrometry imaging (MSI) is an analytical technique that enables the simultaneous detection of hundreds to thousands of chemical species while retaining their spatial information; usually, MSI is applied to biological tissues. Combining these elements can create ion images, which allows for the identification and localization of multiple chemical species within the sample. Being able to produce molecular images of biological tissues has already impacted the study of health and disease; however, the next logical step is being able to combine MSI with quantitative mass spectrometry methods to both quantify and determine the localization of disease progression or drug action. In this tutorial, we will detail the main factors to consider when designing a qMSI experiment and highlight the methods that have been developed to overcome these added complexities, specifically for those newer to the field of MSI.

Infrared matrix-assisted laser desorption electrospray ionization (IR-MALDESI) uses an infrared laser to desorb neutral biomolecules with postionization via ESI at atmospheric pressure. The Gaussian profile of the laser with conventional optics results in the heating of adjacent nonablated tissue due to the energy profile being circular. A diffractive optical element (DOE) was incorporated into the optical train to correct for this disadvantage. The DOE produces a top-hat beam profile and square ablation spots, which have uniform energy distributions. Although beneficial to mass spectrometry imaging (MSI), it is unknown how the DOE affects the ability to perform quantitative MSI (qMSI). In this work, we evaluate the performance of the DOE optical train against our conventional optics to define the potential advantages of the top-hat beam profile. Absolute quantification of glutathione (GSH) was achieved by normalizing the analyte of interest to homoglutathione (hGSH), spotting a dilution series of stable isotope labeled glutathione (SIL-GSH), and analyzing by IR-MALDESI MSI with either the conventional optical train or with the DOE incorporated. Statistical comparison indicates that there was no significant difference between the quantification of GSH by the two optical trains as evidenced by similar calibration curves. Results support that both optical trains can be used for qMSI without a change in the ability to carry out absolute quantification but providing the benefits of the top-hat optical train (i.e., flat energy profile and square ablation spots)-for future qMSI studies.

The collection of profile data is standard practice within the field of mass spectrometry (MS). However, profile data collection often results in large data files that require extensive processing times, especially in mass spectrometry imaging (MSI) studies where thousands of high-resolution scans are recorded. Natively collecting centroid MS data is an alternative that effectively reduces both the resulting file size and the data processing time. Herein, high-resolution accurate mass (HRAM) Orbitrap MSI data on mouse liver tissue sections without automatic gain control (AGC) were natively collected in both profile and centroid modes and compared based on the file size and processing time. Additionally, centroid data were evaluated against the profile data with regard to the spectra integrity, mass measurement accuracy (MMA), and the number of lipid annotations to ensure that centroid data did not compromise the data quality. For both native and postacquisition centroided data, the variation in mass measurement accuracy decreased relative to the profile data collection. Furthermore, centroid data collection increased the number of METASPACE database annotations indicating higher sensitivity and greater accuracy for lipid annotation compared to native profile data collection. Profile MSI data was shown to have a higher likelihood of false positive identifications due to an increased number of data points on either side of the peaks, whereas the same trend was not observed in data collected in native centroid data collection. This publication explores and explains the importance in properly centroiding MSI data, either natively or by adequate centroiding methods, to obtain the most accurate information and come to the best conclusions. These data support that natively collecting centroid data significantly improves MMA to sub-ppm levels without AGC and reduces false positive annotations.

Mass spectrometry imaging (MSI) data visualization relies on heatmaps to show the spatial distribution and measured abundances of molecules within a sample. Nonuniform color gradients such as jet are still commonly used to visualize MSI data, increasing the probability of data misinterpretation and false conclusions. Also, the use of nonuniform color gradients and the combination of hues used in common colormaps make it challenging for people with color vision deficiencies (CVDs) to visualize and accurately interpret data. Here we present best practices for choosing a colormap to accurately display MSI data, improve readability, and accommodate all CVDs. We also provide other resources on the misuse of color in the scientific field and resources on scientifically derived colormaps presented herein.

In mass spectrometry imaging (MSI) applications of infrared matrix-assisted laser desorption electrospray ionization (IR-MALDESI), an exogenous ice layer is the gold standard for an energy-absorbing matrix. However, the formation of the ice matrix requires additional time and instrument hardware, so glycerol was investigated herein as an alternative to the ice matrix to potentially improve spatial resolution and ionization, while decreasing experiment time. Glycerol solutions of varying concentrations were sprayed over top of rat liver tissue sections for analysis by IR-MALDESI and compared to the typical ice matrix condition. Additionally, we tested if combining the ice matrix and glycerol matrix would further improve analyses. Matrix conditions were evaluated by comparing ion abundance of six lipid species, the laser ablation spot diameter, and number of METASPACE annotations. The ion abundances were also normalized to the volume of tissue ablated to correct for lower abundance values due to less ablated tissue. It was observed that utilizing a 50% glycerol matrix without ice provides improved spatial resolution with lipid abundances and annotations comparable to the ice matrix standard, while decreasing the time required to complete an IR-MALDESI tissue imaging experiment.

Although glutathione plays a key role in cancer cell viability and therapy response there is no clear trend in relating the level of this antioxidant to clinical stage, histological grade, or therapy response in patient tumors. The likely reason is that static levels of glutathione are not a good indicator of how a tissue deals with oxidative stress. A better indicator is the functional capacity of the tissue to maintain glutathione levels in response to this stress. However, there are few methods to assess glutathione metabolic function in tissue. We have developed a novel functional mass spectrometry imaging (fMSI) method that can map the variations in the conversion of glycine to glutathione metabolic activity across tumor tissue sections by tracking the fate of three glycine isotopologues administered in a timed sequence to tumor-bearing anesthetized mice. This fMSI method generates multiple time point kinetic data for substrate uptake and glutathione production from each spatial location in the tissue. As expected, the fMSI data shows glutathione metabolic activity varies across the murine 4T1 mammary tumor. Although glutathione levels are highest at the tumor periphery there are regions of high content but low metabolic activity. The timed infusion method also detects variations in delivery of the glycine isotopologues thereby providing a measure of tissue perfusion, including evidence of intermittent perfusion, that contributes to the observed differences in metabolic activity. We believe this new approach will be an asset to linking molecular content to tissue function.