Predicting xenobiotic metabolism: a computational approach mining LC-MS/MS data with SIRIUS and BioTransformer.

Area of Science:

• Pharmacology
• Analytical Chemistry
• Computational Biology

Background:

• Drug biotransformation and bioactivation are crucial in drug discovery.
• Identifying drug metabolites is challenging due to prediction difficulties and the need for prior knowledge.
• Untargeted metabolomics approaches require robust metabolite identification strategies.

Purpose of the Study:

• To develop and validate a computational workflow for predicting and identifying drug metabolite structures.
• To assess the workflow's efficiency in analyzing drug metabolism in human hepatocytes and liver microsomes.
• To accelerate the process of drug metabolite identification and elucidation.

Main Methods:

• Utilized high-resolution LC-MS/MS metabolomics data.
• Integrated BioTransformer and SIRIUS for computational prediction and identification.
• Analyzed metabolites from six model drugs (amitriptyline, carbamazepine, cyclophosphamide, fipronil, phenytoin, verapamil) in human liver systems.

Main Results:

• The computational workflow successfully identified 62-100% of known drug metabolites.
• Four novel metabolite structures were proposed, including one amitriptyline and three verapamil metabolites.
• The approach demonstrated high efficiency in metabolite identification, reducing manual effort.

Conclusions:

• The developed computational workflow effectively automates drug metabolite identification.
• This strategy enhances metabolite coverage and aids in elucidating metabolites of new drugs.
• The workflow shows significant potential for advancing drug metabolism studies in drug discovery.