Fuzzy-Inference System for Isotopic Envelope Identification in Mass Spectrometry Imaging Data

Mass spectrometry is one of the widely used techniques in proteome studies, enabling, inter alia, the identification of proteins present in biological samples based on the analysis of unique peptides originating from the proteins of interest. It should be noted, however, that from the point of view of mass spectrometry data pre-processing, the identification of an isotopic envelope of a peptide plays a crucial role in its precise annotation. Nowadays, there are a plethora of algorithms created to reach this goal, nevertheless, they strongly depend on the type of experimental platform (especially the ionization method: MALDI, ESI, etc.) and are usually dedicated to a specific type of molecules (e.g., lipids or peptides). We propose a unique approach that combines information about the spatial distribution of a molecule across a tissue section sample with a fuzzy-inference system. The mass spectrum was considered as a set of peaks. For each peak, an intensity map was constructed, that presents the spatial distribution of peptide abundance across the tissue sample. The obtained intensity map was further analyzed and the outcome of this analysis was applied to the fuzzy-inference system and to the fuzzy C-means image segmentation method.