August 20-24, 2017
Walter E. Washington Convention Center, Washington, DC, USA
Aug. 23, 2017, 10:05 AM–10:30 AM
An Automated strategy for Targeted and Untargeted Metabolite Identification in Xenobiotic Metabolism
Richard Lee, Vitaly Lashin, Andrey Paramonov, Alexandre Sakarov Seminar: Emerging Mass Spectrometry Trends in Support of Agricultural Research & Development
Meeting Room 15, Renaissance Washington DC Downtown
Abstract: Information gained from metabolism studies plays a critical role to determine the viability of a new chemical entity. The site of biotransformation are recognized through interpretation of mass spectrometric data, which ultimately leads to the elucidation of the biotransformation pathway. In the last 10 -15 years there has been a number of technological breakthroughs in both LC/MS hardware and the software which handles the data from these instruments. In addition, isotope (stable heavy or radio) labeling experiments have also played a critical role in detecting metabolites especially utilized in Agro-chemistry. However, there still lies challenges for structure elucidation of metabolites from the parent structure. In this work, we describe a new automated software strategy for detecting potential metabolites from metabolism studies.
The software strategy, was intended to be a vendor neutral and designed to work with several high resolution mass analyzers. Initially, possible metabolite structures were predicted and generated from an assembly based metabolism model. Metabolites were identified based on their accurate mass and theoretical isotopic distribution calculated from molecular formulae. The workflow also takes advantage of additional information incorporating the UV trace, RAD trace, or by special isotope (for labeled-unlabeled parent) filtering depending on the available information. As part of the spectral interpretation strategy, the algorithm was able to assign fragment ions of the parent structure to its respective MS2 spectra. Structures of metabolites were verified and scores were provided by comparing the assigned fragment pairs and those ion pairs which differ by a specific mass shift. In cases where a discrete structure could not be determined, Markush notations were used, until manual curating was performed to allow for changes to the substructure. Finally, both predicted and unexpected metabolites were combined into a single biotransformation map, where all related mass spectra were associated to each element in the map, and uploaded to a knowledge management system for easy data review. As an added benefit, all peak areas from their respective XICs across the study were tabulated in a summary table and graphically displayed as a stability/kinetic plot.
Potential metabolites were then detected based on the predicted list, and as a complement, a non-targeted unexpected metabolite extraction process was combined into the overall processing routine which employs a fractional mass filter within the component detection algorithm.