Lawrence Berkeley National Laboratory

Mass spectrometry imaging (MSI) has recently emerged as a widely used technique in application areas including pharmaceutical evaluation, ecosystem physiology, high-throughput screening and many others. Mass spectrometry images are made by raster scanning a sample in two dimensions and acquiring mass spectra at each location. Although there are many ways of performing MSI and a diversity of suitable samples, data analysis steps are generally consistent across experiments. These standard analysis steps include access to and management of the raw data, processing of individual spectra with centroiding and denoising algorithms. They also include normalization approaches to both improve spot-to-spot variability and quantification of molecules. Finally, multivariate statistics approaches are used to navigate the complexity of sometimes overwhelming large volume of information. Here, we present all of these analyses approaches and discuss how and why they are used in MSI.