2017 Research Project
Machine Learning Approaches to Predicting Drug Success
The exceptionally high cost and time of developing new drugs is a critical challenge of modern translational research. The ability to predict the eventual success of early drug candidates with greater accuracy could help to reduce this cost and ensure that drugs that are likely to be successful are prioritized. Computational methods to facilitate this prediction have been attempted. This project will develop a novel machine learning-based approach using variables that have largely been ignored but that we believe may have considerable predictive value.
Currently, a pilot study model has been completed which focusses on small molecule therapies. The 2017 SRF-Oxford-CTSCC project will now expand the model to include further indications and parameters. Students participating in this research endeavor will be allocated a section of a database and will be responsible for collecting information that will be used to improve the predictive model. Such a predictive approach may allow identification of promising therapeutics in development and additionally may identify important predictors of eventual drug success for further in-depth investigation.
This project is open for applications from students from any background. Prior research experience is not necessary. However, please be sure to highlight any experience in computational modelling or statistical analysis or familiarity with the drug development process in your cover letter.