AI is also helping to create new medicines

The start-up Insilico Medicine, supported by Nvidia, uses generative artificial intelligence to design new drugs dedicated to so-called rare diseases. Diseases which nevertheless affect some 300 million people in the world, but only 5% of whom have a treatment to cure them.

Thanks to the use of generative AI, the company has managed to generate a new molecule to treat idiopathic pulmonary fibrosis (IPF), a disease suffered by five million people worldwide and whose median survival is estimated between three and four years old. To generate this molecule, Insilico developed an AI platform based on a generative adversarial network and then integrated its algorithms into the Pharma.AI platform.

A major step forward

Following the satisfactory results obtained by phase 1 of its clinical trial, the company announces that it is entering phase 2. “This is the first time that a drug for an AI-discovered and engineered target has reached this stage”, specifies the start-up. This new test phase consists in judging the effectiveness of the molecule INS018_055, already administered orally for twelve weeks to subjects suffering from idiopathic pulmonary fibrosis. The company now plans to recruit, using AI, other subjects with this pathology for more extensive and revealing tests.

Usually, the search for drugs for this type of disease takes a long time and proves terribly expensive for laboratories. However, the arrival of generative artificial intelligence is changing the situation today. It makes it possible to discover and design more quickly the discovery and design of new molecules intended to fight against these rare diseases. Without the help of AI, the research would have cost more than $400 million and would have taken six years. Generative AI allowed Insilico to get there for one-tenth the cost and one-third the time. The company has indeed reached the first phase of clinical trials two and a half years after the start of the project.