Over the past few years, biochemist John Jefferson Perry at the University of California, Riverside, has collaborated on a number of projects with Atomwise Inc., a company that uses artificial intelligence, or AI, for drug discovery. Now Perry and the company have formed a joint venture called Theia Biosciences.

Perry’s collaboration with Atomwise began in 2017 when he received its Artificial Intelligence Molecular Screening Award. He said the collaborative research has been moving forward rapidly since then.

“Theia Biosciences is building off of the initial research started in collaboration with Atomwise during my first AIMS Award,” said Perry, an assistant professor of biochemistry and co-founder of Theia Biosciences. “Theia Biosciences is focused initially on meeting an unmet medical need in helping treat age-related macular degeneration through developing small molecule inhibitors. I am excited to form this partnership to produce novel and very much needed therapies.”

Age-related macular degeneration, or AMD, is the leading cause of blindness in people over the age of 50 in the developed world. AMD affects a part of the retina called the macula and causes a blurring of the sharp, central vision needed for several activities. While AMD does not cause complete blindness, the loss of central vision can make it difficult for a person to function.

AMD has two main types. In the dry form, the macula thins and dries out, eventually losing its function. In the wet form, abnormal blood vessels under the retina begin to grow toward the macula, leading to macula damage and a rapid and severe loss of central vision. Over time, the dry form can turn into the wet form.

Current estimates show that of the approximately 170 million people affected worldwide with AMD, 80%-90% have the dry form, which has no existing therapy. Variants of the HTRA1 gene are known to increase susceptibility to AMD. Theia Biosciences plans to develop small molecule inhibitors of the HTRA1 protein as a potential first-in-class treatment for the wet form and potentially also the dry form of AMD.

“We were excited to have shown proof-of-principle in my lab at UCR, so it seemed like an opportune time to try to move forward with a joint venture,” Perry said. “My lab and Atomwise have very synergistic strengths and skill sets, which can hopefully be leveraged to develop novel small molecule-based therapeutics for the clinic.”

In the research collaboration, small molecule inhibitors of HTRA1 were selected using an AI-based drug discovery technology called AtomNet, a proprietary and deep convolutional neural network developed by Atomwise. This machine-learning approach can rapidly screen billions of chemical compounds in order to discover and optimize small molecule therapies quickly and efficiently, reducing the time from identifying a novel protein target to generation of preclinical drug candidates. 

“An important mission at Atomwise is to use AI-based technology to make better medicines faster,” said Abraham Heifets, CEO and co-founder of Atomwise. “By collaborating with academic investigators like Dr. Perry and UC Riverside, we hope to dramatically increase drug discovery efforts worldwide. We are excited to see Dr. Perry, one of the earliest awardees of our AIMS Awards Program, form Theia Biosciences with Atomwise.”

The formation of Theia Biosciences to develop drugs targeting HTRA1 is the first academic joint venture to arise from a series of collaborations between Perry’s lab and Atomwise. Nine additional projects are underway at Atomwise to identify small molecule candidates for oncology and other age-related disease targets.

The UCR Office of Technology Partnerships assisted Perry and Atomwise on the research agreements. Through these agreements, Perry and Atomwise were able to collaborate using Atomwise’s AI technology to design compounds that would best bind to proteins.

“We hope to continue to use at Theia Biosciences the cutting-edge AI-based tools developed by scientists at Atomwise,” Perry said.

Atomwise invented the first deep-learning AI technology for structure-based small molecule drug discovery. Created in 2012, Atomwise performs hundreds of projects per year in partnership with some of the world's largest pharmaceutical and agrochemical companies, as well as more than 200 universities and hospitals in 40 countries.