AI and drug discovery: a slow evolution in cancer care

Artificial intelligence is being deployed throughout the biopharmaceutical sector by companies trying to bring products to patients faster, and at a lower cost. Hopes that this technology would already be triggering a seismic shift in drug discovery have proved optimistic. But AI is here to stay, and important indicators of progress are set to emerge from some of the leading players in 2025.

The complex world of cancer drug development is the primary focus for many of the AI groups. Powerful computational platforms are being used to find new biological targets, visualize and design new compounds, assess potential toxicities, or screen huge libraries of existing drugs to help discover new uses.

The ultimate aim here is to improve pipeline attrition rates. Most drug research fails, and the cost of setback is substantial. In 2022, Deloitte estimated that on average it costs $2.2bn to bring a new drug to market.

To date, few AI-designed research projects have made it into clinical trials, and much early work has failed, knocking confidence in these approaches. It is too early to measure AI’s impact on R&D productivity, but significant promise remains in utilizing AI to accelerate cancer drug discovery.

So, what is coming next in this space? Important data will emerge from several closely watched projects in the coming months, from the leading names in AI drug discovery.

Schrödinger is a foundational name in AI drug discovery. It started in 1990 as a software company – Microsoft founder Bill Gates was an early investor – evolving to offer high throughput screening, protein structure prediction and molecular discovery and design, licensing these platforms to other developers or spinning the technology out into standalone companies like Nimbus. It took its first wholly owned project into the clinic in 2022 and now has three cancer projects in phase 1, all of which will yield data in 2025.

Relay Therapeutics is a younger company having been founded in 2016, but it has already successfully out licensed its lead oncology project for a substantial sum. A second breast cancer project will yield further data in 2025 and enter a pivotal trial.

Recursion Pharmaceuticals is only a decade old, but it already has six projects in the clinic. Two of these are oncology projects, and they will be joined by two further novel agents next year. Expect data towards the end of 2025 and into 2026.

For now, the power of these platforms likely lies in their ability to design best-in-class approaches to well-validated mechanisms. The unlocking of novel disease targets and pathways was always going to be the big challenge. Data due in the coming months and years will show whether these approaches are bearing fruit.