British startup uses AI to develop new drug compound, in a world-first breakthrough

British startup uses AI to develop new drug compound, in a world-first breakthrough
Innovation

A British startup has used artificial intelligence to create a new drug molecule that will soon enter clinical trials, in an unprecedented breakthrough for the use of machine learning in drug development, according to BBC News.

Designed to treat patients with obsessive-compulsive disorder (OCD), the drug was created in a collaboration between the Oxford AI startup Exscientia and the Japanese pharmaceutical manufacturer Sumitomo Dainippon Pharma. 

While drugs typically take several years to enter into clinical trials, the new drug qualified within 12 months, meeting all the same standards required for phase I clinical trials in Japan. 

Development of new drug treatments costs an average of $2.6 billion, and AI could also make the process much cheaper. 

Faster, less expensive drug development could lead to new treatments for a wide range of conditions. Exscientia is already working on other drugs to treat cancer and cardiovascular diseases.

“I think AI has huge potential to enhance and accelerate drug discovery,” according to Institute of Cancer Research chief executive Paul Workman. “I’m excited to see what I believe is the first example of a new drug now entering human clinical trials, that was created by scientists using AI in a major way to guide and speed up discovery.” 

First, the firm used a set of algorithms to determine the best chemical structure for a molecule to target receptors in the brain involved in OCD. The algorithms generated millions of possible molecular structures, and then filtered the results to select those most suitable to be tested. The new compound is called DSP-1181.

“The AI can learn faster than conventional approaches, so we had to make and test only 350 compounds, a fifth of the normal number of compound candidates, which is record-breaking productivity,” according to Exscientia chief and molecular biologist Andrew Hopkins. 

Hopkins says the same algorithms “can be applied to any drug targets, against a huge range of diseases in oncology, cardiovascular and rare diseases.”

The startup is working with a range of other drug companies, including Bayer, GlaxoSmithKline, and Sanofi. It’s raised $43 million from Bristol-Meyers Squibb, Evotec, and other investors.

Human trials for the OCD compound are set to begin in March, and will take place in Japan. It will mark the first time a drug designed by AI will be used on humans. If successful, the drug will face further testing around the world. 

Image by HeungSoon from Pixabay