Linkedin image/svg+xml Youtube
User

SIRIUS 🤝 BioTransformer: A Dream Team for Unraveling Drug Metabolism

Understanding the metabolic fate of pharmaceutical drugs is a critical component of modern drug discovery and safety testing. Tracking these metabolic changes is notoriously difficult. Identifying the precise structures of drug metabolites within complex mixtures remains a massive analytical bottleneck. To overcome this hurdle, we have integrated BioTransformer in SIRIUS. Researchers have benchmarked this approach using six diverse model drugs and found that it not only bypasses the need for exhaustive manual identification but actively expands our structural coverage of the metabolome.
Understanding the metabolic fate of pharmaceutical drugs is a critical component of modern drug discovery and safety testing. (Image by qimono from Pixabay)

Getting rid of the xenobiotics in your body

When a xenobiotic enters the human body, its journey is highly complex. Approximately 75% of drugs undergo biotransformation to facilitate their elimination1,2. While this process typically ensures safe clearance, it can occasionally lead to bioactivation, producing reactive and potentially toxic metabolites that induce cellular damage3,4,5.

Elucidating xenobiotic metabolism remains a notorious bottleneck in drug discovery and environmental toxicology. While modern liquid chromatography-tandem mass spectrometry (LC-MS/MS) can detect thousands of signals in a single biological sample, putting a structural name to those signals is exceptionally difficult6. Public chemical databases focus heavily on natural, endogenous molecules, leaving xenobiotic transformation products poorly documented. Consequently, researchers have traditionally relied on tedious, manual literature searches, meaning that unexpected or novel metabolites are frequently missed.

The term “xenobiotic” refers to a highly diverse range of foreign compounds that a living system is exposed to, such as pharmaceutical drugs, cosmetics, nutritional phytochemicals, and agrochemicals. When the body is exposed to xenobiotics, its metabolic system works to safely eliminate them through biotransformation, e.g. by adding hydrophilic functional groups to the xenobiotics, making them easier to excrete from the body.

To address this challenge, researchers from the Vrije Universiteit Amsterdam evaluated a way forward, establishing and validating an automated computational workflow that transitions away from manual annotation toward integrated small-molecule identification7.

The Solution: An Automated Workflow Driven by SIRIUS

The research team tested an automated computational workflow combining high-resolution LC-MS/MS metabolomics data with BioTransformer8,9 and SIRIUS to predict and putatively identify drug metabolite structures.

BioTransformer: Instead of querying massive, noisy public chemical repositories, the team used BioTransformer 3.0 to generate focused, drug-specific metabolite databases. BioTransformer logically simulates human metabolism by predicting phase I (via cytochromes P450) and phase II conjugation pathways directly from the parent drug’s structure.

SIRIUS utilizes MS/MS fragmentation trees and isotopic patterns to determine the molecular formula of unknown signals. It then predicts the underlying molecular fingerprint of the compound. By matching the predicted fingerprints against the predicted biotransformations, the workflow prioritizes smarter data over oversized databases, dramatically reducing false positives and streamlining annotation.

Benchmarking Against Manual Annotation

To validate the efficiency of this computational approach, the researcher team benchmarked it against a conventional manual annotation strategy using six diverse model drugs: amitriptyline, carbamazepine, cyclophosphamide, fipronil, phenytoin, and verapamil. These compounds were administered to human liver microsomes (HLMs) and primary human hepatocytes (PHHs)10.

The automated SIRIUS and BioTransformer workflow successfully identified 62% to 100% of the drug metabolites discovered through rigorous manual literature-based tracking. Standard public databases like PubChem and PubMed (PubChem/Med) yielded a massive rate of false positives, returning between 500 and 1,400 suggested structural options per feature. Strikingly, less than 2% of those public database candidates were actually drug-related. Limiting the chemical space to focused BioTransformer candidates yielded higher confidence scores and accelerated structural tracking.

Closer inspection of the few metabolites missed by the automated computational setup revealed that they primarily suffered from low mass spectrometry signal intensities.

  • Amitriptyline: A tricyclic antidepressant that boosts serotonin and norepinephrine. Primarily used at low doses for chronic pain, migraines, and insomnia, as well as major depression.
  • Carbamazepine: An anticonvulsant that calms overactive brain nerves. Used to control epileptic seizures, stabilize bipolar mood swings, and treat trigeminal neuralgia (facial nerve pain).
  • Phenytoin: An anticonvulsant that prevents rapid electrical misfires in the brain. Used specifically to prevent and treat complex partial and tonic-clonic (grand mal) seizures.
  • Verapamil: A calcium channel blocker that relaxes heart muscles and blood vessels. Used to treat high blood pressure, angina (chest pain), and irregular heart rhythms.
  • Cyclophosphamide: A heavy-duty chemo agent and immunosuppressant that halts cancer cell DNA replication. Used to treat leukemia, lymphoma, breast cancer, and severe autoimmune diseases.
  • Fipronil: A potent insecticide (not for human use) that disrupts insect nervous systems. Commonly used in pet flea/tick topicals and agricultural pest control.

Native BioTransformer Integration in SIRIUS 6

Since SIRIUS 6.2, we have natively integrated BioTransformer in the SIRIUS GUI. This development eliminates the need for manual script configurations, making automated transformation product analysis accessible directly within the SIRIUS GUI.

The researcher team performed parallel verification analyses to confirm whether this native BioTransformer integration in SIRIUS matches the accuracy of external BioTransformer database creation. The testing proved that embedded BioTransformer feature yields highly analogous, robust results.

Discovery of Four Novel Drug Metabolites

Beyond merely matching known compounds, the computational workflow demonstrated its predictive power by discovering four completely new, previously unreported drug metabolites that were missed by traditional manual tracking: a novel structural isomer for an amitriptyline metabolite (AM4) and three completely new metabolic products of the blood pressure medication verapamil (VM1, VM4, and VM11).

  • AM4 was discovered as a structural isomer of the well-known metabolite nortriptyline. While sharing an identical molecular formula and highly similar MS/MS fragmentation patterns, AM4 exhibited a distinct retention time and non-interchanging kinetic formation over time.
  • VM1 was proposed with a structure bearing structural similarity to verapamil. While the molecular formula had been noted in in vivo literature, the actual structural connectivity generated de novo by SIRIUS corrected prior literature assumptions based on fragmentation tree data.
  • VM4 was detected clearly in negative ion mode and predicted accurately by BioTransformer. This structure was previously known only as a synthetic chemical product but was identified here for the first time as an active biological metabolite.
  • VM11 was identified in HLM incubations. It shares a molecular formula with known rat metabolites, but its specific structural arrangement was resolved here for the first time via de novo fragmentation analysis.

Why This Matters

The combination of SIRIUS and BioTransformer represents a major step forward for small-molecule analysis, pharmacology, and toxicology by accelerating the identification of unknown degeredation products. Crucially, it empowers scientists to detect potentially toxic, reactive intermediates early in preclinical testing, driving forward the development of safer and more effective pharmaceuticals.

References
  1. F.P. Guengerich. Cytochrome P450, 4th edn. Springer US (2005) https://doi.org/10.1007/0-387-27447-2 ↩︎
  2. L. Di. The role of drug metabolizing enzymes in clearance. Expert Opin. Drug Metab. Toxicol. (2014) https://doi.org/10.1517/17425255.2014.876006 ↩︎
  3. M.P. Holt, C. Ju. Mechanisms of drug-induced liver injury. AAPS J. (2006) https://doi.org/10.1208/aapsj080106 ↩︎
  4. S. David, J.P. Hamilton. Drug-induced liver injury. US Gastroenterol. Hepatol. Rev. (2010) ↩︎
  5. F.P. Guengerich. A history of the roles of cytochrome P450 enzymes in the toxicity of drugs. Toxicol. Res. (2021) https://doi.org/10.1007/s43188-020-00056-z ↩︎
  6. A.C. Schrimpe-Rutledge, S.G. Codreanu, S.D. Sherrod, J.A. McLean. Untargeted metabolomics strategies-challenges and emerging directions. J. Am. Soc. Mass Spectrom. (2016) https://doi.org/10.1007/s13361-016-1469-y ↩︎
  7. V. Pozo Garcia, M. Zhang, T.S. Çobanoğlu, K. Holm, P. Jennings, D.A. Poole, S. Moco. Predicting xenobiotic metabolism: a computational approach mining LC-MS/MS data with SIRIUS and BioTransformer. Arch. Toxicol. (2026) https://doi.org/10.1007/s00204-025-04248-0 ↩︎
  8. Y. Djoumbou-Feunang, J. Fiamoncini, A. Gil-de-la-Fuente. BioTransformer: a comprehensive computational tool for small molecule metabolism prediction and metabolite identification. J. Cheminform. (2019) https://doi.org/10.1186/s13321-018-0324-5 ↩︎
  9. D.S. Wishart, S. Tian, D. Allen. BioTransformer 3.0-a web server for accurately predicting metabolic transformation products. Nucleic Acids Res. (2022) https://doi.org/10.1093/nar/gkac313 ↩︎
  10. V. Pozo Garcia, T.S. Çobanoğlu, H.S. Hammer. Nutrient environment improves drug metabolic activity in human iPSCderived hepatocytes and HepG2. Arch. Toxicol. (2025) https://doi.org/10.1007/s00204-025-04139-4 ↩︎

Try SIRIUS now!
All SIRIUS Features

The easy way to comprehensive structure elucidation​

SIRIUS is the comprehensive software solution for the high-throughput identification of small molecules from fragmentation mass spectrometry data. SIRIUS provides a comprehensive set of features spanning every step from feature detection to detailed result validation. It is designed to not only accurately characterize known compounds but also to confidently identify “unknown unknowns” in complex biological samples. 

Discover more

Share