A variety of mass spectrometers are commercially available. Each of these mass spectrometers stores raw mass spectra in a proprietary format. The raw spectra have to be transformed into common data format first. As in any study of biological phenomena, it is crucial that only relevant observations are identified and related to each other. The interpretation and comprehension of the collection of mass spectra presents major challenges and involve several data mining steps. The aim of mass informatics is to reduce data dimensionality and to extract relevant knowledge from thousands of mass spectra (Arneberg, 2007). Figure 1 shows an overall framework for mass informatics in differential proteomics. Most of the components of this framework will be discussed in this paper with the exception of the pathway modeling.
Information flow chart in differential proteomics