Host: The Japanese Society of Toxicology
Name : The 47th Annual Meeting of the Japanese Society of Toxicology
Date : 2020 -
Identification of chemical compounds, such as toxic compounds and their metabolites, is of vital importance for toxicological studies. Mass spectrometry has been widely used as one of the most versatile analytical tools. On the other hand, identifying the chemical structures of compounds in complex sample containing a large number of unknown compounds is still quite difficult to accomplish.
Mass spectral similarity networking, also called as molecular networking, is an approach to organize large amount of fragment spectra according to their spectral similarities. We have developed a data analysis framework employing mass spectral similarity network, to organize, classify, structurally annotate and visualize the complex mass spectral dataset containing unknown compounds. Fragment spectra from the sample of interest are processed and organized as a network on base layer. Reference fragment spectra are acquired from public spectral library, which covers great variety of chemical species, and used to create “reference network”. Further, reference network is divided into user-defined sublayers according to chemical/biochemical/toxicological properties.
The multi-layer representation of mass spectral similarity networks effectively visualizes and categorizes the spectra of unknown compounds in sample. Structural information systematically extracted from reference spectral networks provides structural insights into unknown compounds. This versatile method can be applied to spectral dataset containing a wide range of unknown compounds in toxicological studies.
