BioPharma, Artificial Intelligence

Drug-discovery firm nets $14.5M in Series A funding

X-37 leverages AI technology similar to facial recognition to figure out which small molecules can bind most effectively with targeted enzymes.

Firm leverages AI technology similar to facial recognition to figure out which small molecules can bind most effectively with targeted enzymes.

By Joel Berg

Does the future of drug development lie in a kind of facial-recognition technology for enzymes?

That is the hope of X-37 LLC, a drug-development startup that is using artificial intelligence and a deep neural network developed by San Francisco-based Atomwise.

“We think that this is an approach and a technology that is really going to transform drug discovery across the board,” said Dr. David Collier, CEO of X-37, which is partly owned by Atomwise and is based in South San Francisco, California. It was founded last year.

To expedite that approach, X-37 announced Nov. 14 that it had raised $14.5 million in Series A financing, which will go toward expanding its number of drug development programs, Collier said in a phone interview. The company’s goal is to identify drug candidates that could be brought to market and start human trials by 2022. It employs eight people but expects to hire more over the next year.

The company is targeting enzymes that are recognized as being important in specific diseases, but which pharmaceutical companies have had trouble targeting with small molecules, Collier said.

An example is PIM3, an enzyme that plays a role in cancer. However, there are two similarly structured enzymes, PIM1 and PIM2, Collier said. The idea is to block PIM3 without affecting the other two, which could lead to unwanted side effects.

To figure out which small molecules can block an enzyme, X-37 uses the Atomwise technology to scrutinize the structure of an enzyme and then screen virtual libraries containing millions of small molecules to see which would be good candidates for binding with the enzyme, Collier said. He described the technology as akin to facial recognition.

“Instead of recognizing bumps and curves on faces, you are recognizing structures on pockets of enzymes you want to target,” Collier said.

X-37 takes 100 to 300 of the top candidates, has them made and starts testing them. Researchers feed what they discover back into the Atomwise network so it can learn from the results, Collier said.

“This is incredibly important for discovering new drugs,” said Collier, who also is CEO of Velocity Pharmaceutical Development. Other members of X-37 also come from Velocity.

Most drug discovery companies use artificial intelligence to scan data like patient records, Collier said. Atomwise, by contrast, is one of a few looking at the structures of enzymes. Collier feels Atomwise is a leader in the space.

“They’ve been able to successfully demonstrate that they can do this,” he said, citing the company’s work for other pharmaceutical partners.

X-37’s Series A funding round was led by DCVC Bio and it was joined by Alpha Intelligence Capital and Hemi Ventures. Kiersten Stead and Antoine Blondeau of Alpha Intelligence Capital have joined X-37’s board of directors.

“We are very pleased to be investors in X-37,” Stead said in a statement. “We have watched with great excitement as Atomwise has refined its deep neural network and chemistry expertise. We identified the team of drug developers at Velocity Pharmaceutical Development as ideal partners for Atomwise because of their experience in target selection and development of drug leads beyond the initial discovery stage. The progress of the combined teams at X-37 has been superb.”

The company takes its name from an aircraft being used by the U.S. Air Force for sub-orbital missions, Collier said. Velocity named its spin-off drug companies after World War II fighter planes, such as the Corsair, Mustang and Thunderbolt.

“Part of the theme of all of that is speed, which fits with our philosophy at Velocity of trying to streamline drug development,” said Collier, who also is a licensed pilot. The founders of X-37 decided to name the company after the fastest airplane they knew of.

X-37 is structured as an LLC and plans to create subsidiary companies based on the individual drugs it develops. The structure will allow it to sell those companies and the underlying drugs without having to sell the parent company, Collier said. It also offers tax benefits, since profits and capital gains in an LLC are not taxed at the corporate level and then again after they are distributed as dividends. They are taxed at the individual shareholder level only.

Photo: Abscent84, Getty Images