Translational medicine is an emerging interdisciplinary field that seeks to bridge the gap between basic science research and clinical medicine. Many candidate drugs fail in human clinical trials because they are found to be unsafe or ineffective, despite promising nonclinical studies in animal and cell models. Machine learning enables life science organisations to extract more insight from their available data to increase the success rates of their therapeutics, while reducing the testing burden. It can build models that predict efficacy and toxicity and help to identify useful biomarkers. But machine learning can be limited by the size and sparse, noisy nature of many pharmaceutical datasets. The Alchemite technology overcomes these limitations.
Case studies
Predicting pharmacokinetics parameters and curves
In work published in Molecular Pharmaceutics, the Alchemite™ method was applied to the prediction of PK parameters, based on compound structure and sparse in vitro data from AstraZeneca. This work can reduce time, cost, and number of animal studies in late-stage drug discovery.
Modelling toxicology data
Researchers from JTI and Intellegens published a machine learning study of toxicology in the Journal of Chemical Information and Modeling. Analysing experimental datasets to understand and enable control of the toxicological properties of chemicals is a key task in life sciences R&D. The performance of imputation-based machine learning methods was compared to well-established ML-based QSAR methods, finding a significant improvement in the quality of results.
Machine learning for precision medicine
Usually, limitations in available data and computational power render the task of predicting therapeutic efficacy from clinical trials and animal studies information highly challenging. Intellegens and the A*STAR Institute applied Alchemite™ to identify the best dosage of a stem cell therapy treatment to use in treating specific patients for cartilage damage. Better-targeted treatment improves patients’ chances of successful recovery.
Alchemite™ for translational medicine
Alchemite™ technology enables you to:
- Capture complex property-to-property and structure-to-property relationships
- Build models that can predict ADME and other properties
- Use accurate uncertainty estimates to prioritise compounds most likely to succeed
- Focus your experimental resources on the most valuable measurements
- Explore chemical space beyond conventional QSAR models
- Standardise and share data and models while ensuring rigorous security for your IP
