Optibrium’s New Approach Predicts Drug Metabolism Pathways with High Accuracy
Optibrium, a leader in drug discovery software and AI solutions, announced the publication of a groundbreaking study in Xenobiotica titled “Predicting routes of phase I and II metabolism based on quantum mechanics and machine learning”. This research presents a revolutionary method for identifying the metabolic pathways and metabolites of drug candidates early in the drug discovery process.
Unexpected drug metabolism can lead to the failure of promising candidates at late stages of development, even requiring the withdrawal of already approved drugs. Therefore, early computational prediction of the dominant metabolic pathways is crucial for increasing a drug’s chance of success.
The study focuses on the development and validation of Optibrium’s WhichEnzymeTM model, which accurately predicts the enzyme families most likely to metabolize a drug candidate. This new model is combined with Optibrium’s previously published models:
- Regioselectivity models for key Phase I and Phase II drug metabolizing enzymes, using quantum mechanical simulations and machine learning to predict sites of metabolism and resulting metabolites.
- WhichP450 model, predicting the specific Cytochrome P450 isoform(s) responsible for a compound’s metabolism.
By analyzing the combined model outputs, Optibrium developed a novel method for determining the most likely routes of metabolism and metabolites observed experimentally. This method demonstrates high sensitivity in identifying experimentally reported metabolites, exceeding the precision of other existing methods for predicting in vivo metabolite profiles. This enables researchers to identify compounds with greater metabolic stability and improved safety profiles. This innovative approach forms the foundation of Optibrium’s recently launched StarDrop Metabolism module.
Dr. Mario Öeren, principal scientist at Optibrium, emphasizes the significance of the research:
“Our latest study is the culmination of six years of focused research, culminating in a practical model that enables users to predict metabolic pathways for a diverse range of drug-like compounds. Through our carefully curated datasets and unique reactivity-accessibility approach, we built accurate isoform-specific regioselectivity models for crucial Phase I and II enzyme families.”
“Furthermore, using the same data, we trained models that predict the likely enzyme families and isoforms responsible for a compound’s metabolism. Ultimately, by combining these models, we developed and validated a ‘model of models’ that predicts the full metabolic pathway for any given compound,” Dr. Öeren adds.
Continue to check our website soundhealthandlastingwealth.com for more articles of this kind. And, please use our comment section as well, we would love to hear from you.
