Published December 12, 2022 | Version v1
Presentation Open

Exploring the potential for a fragmentation alignment index for database matching

  • 1. University of Amsterdam
  • 2. University of Queensland

Description

Non-targeted analysis using high-resolution mass spectrometry is becoming a critical tool for identifying small molecules of interest across multiple areas of natural and life sciences. The rapid advancement in high resolution mass spectrometry has allowed the collection and archiving of accurate mass spectra of small molecules from hundreds of thousands of biological and environmental samples. To facilitate rapid identification of small molecules, in recent years publicly available online databases of mass spectrometry reference spectra have been built allowing spectral matching (i.e., library searching). However, the fragmentation pattern, including the fragment masses and relative intensities, of a molecule are influenced by experimental factors such as collision energy, collision gas, dissociation technique, instrument type, and ionization mode. Hence, this all impacts the spectral matching.

To assess and potentially overcome this issue, we explored the potential of a fragmentation alignment index (FAI ) for small molecules (i.e., mass < 1000 Dalton) obtained from MassBank EU. Firstly, all spectra of a single chemical were compared to investigate the influence of manufacturer and collision energy. Other factors, such as temperature, could also be of interest, however, only few parameters are recorded in the database.  Secondly, we investigated the optimal parameters for each instrument type to use for generation of reference spectra to minimize the impact of the experimental conditions on the quality of the spectra. The generated optimal spectra will enable the generation of global spectral such as MassBank EU, that can be used for the structural elucidation of small moleculs. Finally, we evaluate the possibility of developing a fragmentation model for normalisation of mass spectral data to improve spectral matching.

Files

1215 Denice van Herwerden.pdf

Files (2.6 MB)

Name Size Download all
md5:9e079ed6cdd0e4c9cb1eeabcefce29bd
2.6 MB Preview Download