How AI is helping scientists discover new drugs

Getty Images

A few patients suffering from a rare disease are currently testing a new drug candidate based on a compound discovered using artificial intelligence. This new technology can save scientists an enormous amount of time compared to conventional research techniques.

A start-up called Insilico Medicine, supported by Nvidia, is using generative artificial intelligence to design new drugs for rare diseases. These kinds of diseases affect some 300 million people worldwide, but only 5% have a cure.

Thanks to the use of generative AI, the company has succeeded in identifying a new compound that could treat idiopathic pulmonary fibrosis (IPF), a disease affecting five million people worldwide, and with median survival estimated at between three and four years.

To identify this compound, Insilico developed an AI platform based on a generative adversarial network, then integrated its algorithms into its Pharma.AI platform.

Following the satisfactory results obtained in phase 1 of its clinical trial, the company has announced that it is entering phase 2.

According to the start-up, this is the first time that a drug aimed at a specific target, discovered and designed by AI, has reached this stage. This new phase of testing involves assessing the efficacy of the INS018_055 compound, already administered orally for 12 weeks to subjects suffering from idiopathic pulmonary fibrosis.

The company now plans to use AI to recruit other subjects suffering from this pathology for more extensive and informative tests.

Traditionally, the search for drugs for this type of disease has been extremely time-consuming and costly for laboratories.

But the arrival of generative artificial intelligence is changing the game. It has made it possible to discover and design new compounds to combat these rare diseases more rapidly.

In this case, without the help of AI, research would have cost over $400 million and taken six years.

Generative AI enabled Insilico to achieve this at one-tenth the cost and in one-third of the time. In fact, the company reached the first phase of clinical trials two and a half years after the start of the project.