Spatial Probabilistic Mapping of Metabolite Ensembles in Mass Spectrometry Imaging

Abstract

Mass spectrometry imaging (MSI) vows to enable simultaneous spatially-resolved investigation of hundreds of metabolites in tissue sections, but it still relies on poorly defined ion images for data interpretation. Here, we outline moleculaR, a computational framework in R, that introduces systematic probabilistic mapping and point-for-point statistical testing of metabolites in tissue to MSI. Beyond statistics, moleculaR allows for arithmetic operations within the same MS image and thereby, for instance, analysis and visualization of complex scores like the adenylate energy charge ([ATP]+0.5*[ADP])/ ([ATP]+[ADP]+[AMP]). moleculaR also enables collective molecular projections, for example of all potassium versus all sodium adducts for spatially-resolved investigation of ion milieus, or for surveys of lipid pathways or other user-defined biomolecular ensembles.