University of Toronto and Insilico Medicine Announce Breakthrough in Quantum AI for Drug Discovery
A recent study led by researchers from the University of Toronto and Insilico Medicine has demonstrated the potential of quantum computing and artificial intelligence (AI) in transforming the drug discovery process.
This research marks the first instance of successfully using quantum computing and AI to identify molecules capable of interacting with biological targets.
The study utilised a combination of quantum computing, generative AI, and traditional computing methods to develop molecules targeting a protein known as KRAS. This protein, associated with uncontrolled cell growth, is a major driver of cancer and has long been considered difficult to target with drugs.
Although mutations in KRAS are present in nearly 25 per cent of human cancers, only two FDA-approved drugs currently exist, offering limited survival benefits compared to conventional chemotherapy.
This approach has the potential to significantly reduce the time required for the early stages of drug discovery, traditionally reliant on screening extensive chemical libraries. Such conventional methods are expensive, time-consuming, and require substantial physical resources. In contrast, computational approaches allow for large-scale virtual screening without the need for physical storage or robotic screening processes.
While the study confirms that quantum computing can play a role in drug discovery, it does not yet establish a definitive advantage over classical computing in this field.
However, as quantum computing technology advances, it is expected to contribute more effectively to AI-driven drug discovery methods.
Following the success of this study, the research team is now applying the hybrid quantum-classical model to other challenging protein targets. Many of these proteins, like KRAS, are small and lack the surface features required for drugs to bind easily.
Additionally, efforts are underway to further refine the two most promising KRAS-targeting compounds, with the aim of advancing them into preclinical testing using animal models.
