Data preprocessing, visualization, and statistical analyses of nontargeted peptidomics data from MALDI-MS

Abstract

Mass spectrometric (MS) comparative analysis of peptides in biological specimens (nontargeted peptidomics) can result in large amounts of data due to chromatographic separation of a multitude of samples and subsequent MS analysis of numerous chromatographic fractions. Efficient yet effective strategies are needed to obtain relevant information. Combining visual and numerical data analysis offers a suitable approach to retrieve information and to filter data for significant differences as targets for succeeding MS/MS identifications. Visual analysis allows assessing features within a spatial context. Specific patterns are easily recognizable by the human eye. For example, derivatives representing modified forms of signals present are easily identifiable due to an apparent shift in mass and chromatographic retention times. On the other hand numerical data analysis offers the possibility to optimize spectra and to perform high-throughput calculations. A useful tool for such calculations is R, a freely available language and environment for statistical computing. R can be extended via packages to enable functionalities like mzML (open mass spectrometric data format) import and processing. R is capable of parallel processing enabling faster computation using the power of multicore systems. The combination and interplay of both approaches allows evaluating the data in a holistic way, thus helping the researcher to better understand data and experimental outcomes.