Mass Spectrometry Data Mining Computational Strategies and Methods for Compound Identification

The latest technical advances in mass spectrornetry (MS) enable the mapping of complex biological samples in a single analvtical run. The possibility to investigate organisms in this manner opens up new possibilities, especially in life sciences. Research fields, such os metabolemics and lipidomics heavily utilize modern chromotography-MS hyphenations for metabolome or lipidome profiling. respectively. In addition, data-dependent information can be accessed in parallel, due to the Iatest hybrid mass spectrometers capabilities. Hence, structural elucidation can be performed to some extent for the most abundant compounds detected. Manual data interpretation is very common. Especially in larger projects, where thousands of samples are analyzed, this can be very tedious and time-consuming, while minor but significant compounds are likely missed. Accordingly, the interpretation and evaluation of resulting multidimensional data sets are the bottleneck regarding overall analysis times. To reduce analysis times, while simultaneously deriving as much as possible knowledge from the data sets, research groups and instrument vendors have developed various software solutions for MS data mining. Thus, in this thesis one of the most popular open-source MS dato mining software, namely MZmine 2, was utilized, enhanced, and extended to face various analytical challenges.