Silicon Valley-based nonprofit scientific research institute SRI International said today it’s spinning out its patented Synfini drug discovery platform as an independent startup named Synfini Inc.

Synfini combines technologies that include physical and virtual chemistry, neuro-symbolic artificial intelligence, a vast proprietary chemistry dataset and robotic automation to develop novel molecular compounds as promising drug candidates.

The platform was developed by SRI during a $40 million, multiyear development effort that was partly funded by the U.S. Defense Advanced Research Projects Agency. SRI, which is more than 75 years old, is known to conduct secret research on behalf of numerous U.S. government agencies and is hugely influential within Silicon Valley, having close ties with leading venture capital firms and also many of its largest companies.

SRI has spun out dozens of startups and licensed key technologies to global corporations, creating a tremendous amount of wealth that is shared among its researchers thanks to its unique nonprofit business model. The impressive catalog of technologies SRI has given birth to include the Department of Defense’s Advanced Research Projects Agency Network, known as ARPANET, which eventually morphed into the public internet, wireless networks, ultrasound and other innovations. It has developed groundbreaking applications in numerous industries over the years, including robotics, education, satellite technology, medical devices and, of course, drug discovery.

Synfini promises to make its presence felt in the latter industry. Its founding team is comprised of scientists and engineers, who possess a combined 150 years of experience in synthetic and medicinal chemistry, engineering and startup expertise.

The Synfini platform is made up of multiple software and hardware-based solutions developed at SRI. It’s said to include a neuro-symbolic AI molecule designer, a computational synthetic planning tool, a high-throughput automated chemistry system and a flow chemistry hardware platform that performs reliable multi-step synthesis of molecular compounds.

Synfini’s founding CEO Doug Donzelli told SiliconANGLE that neuro-symbolic AI is a new and evolving subset of AI that combines neural networks with knowledge-guided symbolic approaches to  advance both AI algorithms, such as abstraction, analogy and reasoning, and applications around explainable and safety-constrained decision making. “At Synfini, neuro-symbolic AI accelerates the generation of molecular compound leads and their optimization for use by at least 50% and with at least a 40% reduction in research and development costs,” he said.

Synfini intends to train its neuro-symbolic AI models using its vast proprietary dataset, which has been gleaned from SRI’s research and validation work over the course of the seven years in which the Synfini platform was under development, Donzelli said. “This data is part of the IP that is now owned by Synfini, Inc.,” he explained. “[The database contains] a large pipeline of compounds that are ready to go to the next phase of drug discovery for oncology and human longevity solutions. This pipeline is valuable to drug hunters and big pharma.”

SRI Chief Executive David Parekh said his organization is always excited when the technologies it develops launch on the commercial market. “Synfini’s suite of hardware solutions, combined with cutting-edge AI technologies, holds great promise for accelerating the speed at which biopharma companies can develop new therapeutics,” he said.

Synfini said it’s targeting pharmaceutical companies with its platform, and says it will help reduce both the time and cost of drug discovery. It said its primary role is to accelerate molecular lead identification and optimization in order to reduce the time it takes for researchers to discover preclinical trial drug candidates.

Researchers will be able to understand the potential capabilities and application of new drug candidates much faster, Synfini said. Its platform can also help pharmaceutical companies to streamline the go-to-market process for new drugs while reducing the cost of their research.

Image: Freepik