AI for pharma and research
- Many target identification programmes that fail for efficacy reasons showing poor association between the drug target and the disease
- We used a semi-supervised classification approach to explore whether gene – disease association data from the Open Targets platform is sufficient to predict therapeutic targets that are actively being pursued by pharmaceutical companies or are already on the marke.
- We show that a neural network is able to predict therapeutic targets with over 70% accuracy demonstrating that disease association is predictive of the ability of a gene or a protein to work as a drug target
Enrico Ferrero, Scientific Leader, Computational Biology, GSK