How Artificial Intelligence Matches Drugs to Patients

Sitting in front of a microscope, Dr. Talia Cohen Solal examines human brain cells grown in petri dishes.

“The brain is a very delicate, complex and beautiful thing,” she says.

Dr. Cohen Solal, a neuroscientist, is the co-founder and CEO of Israeli healthtech company Genetika+.

Founded in 2018, the company says its technology can optimally match antidepressants to patients, avoid unwanted side effects and ensure that prescribed drugs work as well as possible. increase.

“For the first time, we can characterize the right drug for each patient,” adds Dr. Cohen Solal.

Genetika+ accomplishes this by combining the latest stem cell technology (proliferation of specific human cells) with artificial intelligence (AI) software.

Its technicians can generate brain cells from a patient’s blood sample. These are then exposed to several antidepressants, and cellular changes called “biomarkers” are recorded.

This information, along with the patient’s medical history and genetic data, is processed by an AI system to determine the optimal drugs and dosages for doctors to prescribe.

The technology is currently still in the development stage, but Tel Aviv-based Genetika+ plans to commercialize it next year.

As an example of how AI is increasingly being used in the pharmaceutical sector, the company has secured funding from the European Union’s European Research Council and European Innovation Council. Genetika+ also works with pharmaceutical companies to develop new precision medicines.

“We are at the right time to be able to combine the latest advances in computer technology and biological technology,” says Dr. Cohen Solral.

A senior lecturer in biomedical AI and data science at King’s College London, she says AI has so far been used “from identifying potential target genes to treat specific diseases, to discovering new drugs, predicting patient It has helped with everything from improving treatment.” Optimal therapeutic strategies, discovery of biomarkers for personalized patient care, and even prevention of disease through early detection of onset signs.”

The New Tech Economy is a series that explores how technological innovation is shaping the emerging economic landscape.

But fellow AI expert Callum Chase says the adoption of AI across the pharmaceutical sector remains a “slow process.”

“Pharmaceutical companies are huge, and any big change in how they do research and development will impact a lot of people in different departments,” says Chace, author of many books on AI.

“It’s hard to get all these people to buy into the dramatically new way of doing things, partly because older people got to where they are by doing things the old way.

“They know it, they trust it, and they may fear that if the method they know suddenly loses value, it will be less valuable to the company.”

However, Dr. Sailem stresses that the pharmaceutical industry should not rush to compete with AI and take drastic measures before relying on AI predictions.

“AI models can learn the right answers for the wrong reasons. Especially when trained on patient data, it is up to researchers to ensure that various measures are employed to avoid bias. And it’s the developer’s responsibility,” she says.

Hong Kong-based Insilico Medicine uses AI to accelerate drug discovery.

“Our AI platform can identify existing drugs that can be reused, design new drugs against known disease targets, or find entirely new targets and design entirely new molecules.” said co-founder and CEO Alex Zhavoronkov.

Its most developed drug, a drug for a lung disease called idiopathic pulmonary fibrosis, is currently in clinical trials.

Zhavoronkov said that it usually takes four years for a new drug to reach that stage, but thanks to AI, Insilico Medicine got there “within 18 months at a fraction of the cost.”

He adds that the company has another 31 drugs in various stages of development.

Dr. Cohen Solal, who has returned to Israel, says AI can help “solve the mystery” of which drugs work.