Artificial intelligence becomes a tool to manufacture medicines

The possibility of creating a drug for dengue thanks to AI (Artificial Intelligence) is not science fiction and a European NGO has recently launched this project, an initiative that reflects that this technology has a place in the field of medicine.

The NGO “Drugs for Neglected Diseases Initiative” seeks treatments for diseases that generate little interest and in April launched a partnership with BenevolentAI, a British company that seeks to develop molecules thanks to AI.

In early 2020, Scottish company Exscientia, in partnership with Japanese pharmaceutical Sumitomo Dainippon, developed the first molecule “built” thanks to AI that entered clinical trial.

“It’s not futuristic: artificial intelligence is a methodological approach to data processing that can be used at various stages of the pharmaceutical industry’s development process,” says Dr. Thomas Borel, director of scientific affairs at Leem (Federation of Medicines Companies in France)

Visiting the Parisian facilities of French startup Iktos, founded in 2016, he makes it clear that an era is changing. In this company there are no microscopes or traditional biology devices or any white coat employee.

Instead, computers work tirelessly analyzing data at a speed unattainable by any human being.

“The idea is to use existing data to obtain interesting new molecules, faster,” explains Yann Gaston-Mathé, director of the company he co-founded.

For this, his team used a global database with information on one hundred million molecules. From these data, “we create a model that will automatically generate new molecules and identify those that may be active for biological targets of interest,” she explained.

Iktos has a molecule research platform thanks to artificial intelligence that delivers data to pharmaceutical companies subscribed to this service.

Interest of the laboratories

Aqemia, a young company that emerged from the Escola Superior Nacional PSL in 2019 in France, has developed a platform to discover medicines thanks to quantum-inspired statistical physics.

“We use an artificial intelligence that we qualify as generative”, highlights its founder, researcher Maximilien Levesque.

“We invent molecules that will adhere to a specific biological target that is causing a disease. Artificial intelligence feeds on physics: we need to know the physical nature of the molecule and the target to calculate its affinity”, he described.

While startups are at the forefront, labs are increasingly interested in this area. Proof of this are its investments in this sector.

US giant Bristol-Myers Squibb reached a deal with Exscientia last year and could give it more than $1 billion.

In 2019, Swiss lab Novartis and computer giant Microsoft announced a collaboration agreement.

But that doesn’t necessarily mean the end of laboratory chemistry. This new field faces significant difficulties, such as access to data that can be explored and the need to find specialists who, on the one hand, have mastered artificial intelligence and, on the other hand, have knowledge of pharmacology.

There is also a very important regulatory aspect, adds Leem’s Thomas Borel.

“For a drug to be accepted, regulatory systems need to recognize the value of the algorithm,” he says.

“Drugs have been designed with the help of computers for years”, comments Gaston-Mathé, for whom the aim is to provide “additional tools to chemicals without wanting to replace man with machine”.