Comprehensive Mass Spectrometric Mapping of Chemical Compounds for the Development of Algorithms for Machine Learning and Artificial Intelligence

Fully Automated Unconstrained Analysis of High-Resolution Mass Spectrometry Data with Machine Learning

“…10,11 Unfortunately, the main strengths of the method (sensitivity and resolution) result in very complex spectra, drastically increasing the interpretation time and requiring high qualification for the scientist. 12,13 As an example, less than 20% of metabolites were identified in the untargeted analysis. 14,15 It is not uncommon when ∼80% of signals remain uninterpreted in the recorded mass spectrum of reaction mixtures, biomolecular systems, natural organic matter, and others.…”

“…For example, a simple metal complex involved in catalysis Pd 2 Cl 2 (PhBr)-(PhC 2 H) gives a pattern with 17 signals, while the well-known peptide lysozyme gives at least 18 signals for the +9 ion. 12 Finally, the third source is significant in electrospray ionization mass spectrometry-based proteomics, as large molecules tend to form multiply charged species. The deconvolution of multiply charged species was studied in depth.…”

“…Complex mixtures are important in many areas of life sciences and chemistry, from the analysis of natural products , and metabolomics , to the investigation of chemical reaction mechanisms. − Due to its superior resolution and sensitivity, high-resolution mass spectrometry (MS) with an electrospray ionization source is one of the most versatile analysis methods. ,, Modern instruments detect compounds with tiny concentrations down to trace signals and can perform structure elucidation via tandem mass spectrometry experiments. This opens new ways for an in-depth understanding of complex dynamic processes, for instance, in biology and medicine. , Unfortunately, the main strengths of the method (sensitivity and resolution) result in very complex spectra, drastically increasing the interpretation time and requiring high qualification for the scientist. , As an example, less than 20% of metabolites were identified in the untargeted analysis. , It is not uncommon when ∼80% of signals remain uninterpreted in the recorded mass spectrum of reaction mixtures, biomolecular systems, natural organic matter, and others.…”

“…Second, a significant degree of complexity comes from the appearance of “patterns” (groups of many signals) instead of single signals for individual molecules, the so-called “isotopic distributions”. For example, a simple metal complex involved in catalysis Pd 2 Cl 2 (PhBr)­(PhC 2 H) gives a pattern with 17 signals, while the well-known peptide lysozyme gives at least 18 signals for the +9 ion . Finally, the third source is significant in electrospray ionization mass spectrometry-based proteomics, as large molecules tend to form multiply charged species.…”