Data Study Group Final Report: Ignota Labs - Toxicity Prediction for Drug Discovery

Data Study Groups are week-long events at The Alan Turing Institute bringing together some of the country’s top talent from data science, artificial intelligence, and wider fields, to analyse real-world data science challenges.

The Turing DSG Report Dec 2023 by Ignota Labs presents a comprehensive exploration of machine learning’s (ML) capabilities in advancing drug discovery, specifically through the prediction of cardiac
toxicity. The report articulates the persistent challenges in the pharmaceutical domain, including escalating drug discovery costs and declining clinical success rates, which amplify the financial strain on healthcare systems and the pharmaceutical industry. Approximately 90% of clinical-stage drugs fail, underscoring the critical need for improving early-stage prediction of small molecule properties to mitigate late-stage failures, enhance research and development efficiency, lower drug costs,
and expedite novel therapeutics’ delivery to patients.

The initiative by Ignota Labs, in collaboration with the Alan Turing Institute, embarks on a challenge leveraging ML to predict cardiac toxicity by assessing molecular binding to key cardiac ion channels. The aim is to harness advanced ML techniques like Graph Neural Networks (GNNs), multi-task learning, and transfer learning for drug toxicity prediction, while also valuing traditional methods like tabular data representation and ensemble strategies. This holistic approach endeavours to comprehensively address the complex challenges of drug toxicity prediction.

Data Study Group - December 2023 | The Alan Turing Institute