Biotechnology research continues to churn out new drugs addressing genetic drivers of cancer. While that progress gives patients more treatment options, one limitation of this approach is it takes for granted that we know what target to drug, according to Ron Alfa, CEO of startup Noetik.
The problem becomes more glaring with the rise of immunotherapies. If a drug targets a known cancer-driving mutation such as KRAS or EGFR, a clinical trial can be designed to enroll patients with those tumor mutations, Alfa said. But when it comes to immunotherapies, medicine doesn’t have the same understanding of which patients should respond to treatment.
“We need a new tool that allows us to see these tumors in a different way, in a way that allows us to see the underlying tumor biology,” Alfa said.
Noetik might have that tool. Alfa co-founded the startup with Chief Scientific Officer Jacob Rinaldi. Both are veterans of artificial intelligence drug discovery company Recursion. South San Francisco-based Noetik, whose name is derived from the Greek word for “intellectual,” emerged from stealth on Thursday, backed by $14 million in seed financing.
The research of Noetik starts with tumor samples from patients. Those samples are turned into slides—images that are analyzed by the software. The ultimate goal is to find new cancer targets. But Alfa says the approach is different than the tack taken by some in drug research. The first generation of AI drug discovery relied on humans to tell the AI model where to look, Alfa said. It’s called supervised learning. By contrast, Noetik allows the AI technology to learn directly from the data without being instructed what to look at or what to look for. This approach is called self-supervised learning.
“Our hope is that these models pick up things that the models think are the most relevant and are most biologically relevant, things humans might not think of,” Alfa said.
Alfa worked as head of research at Recursion. Rinaldi was head of oncology. Before Recursion, Rinaldi worked in cancer research at Genentech. The executive team also includes Chief Technical Officer Lacey Padron, who previously worked as the vice president of informatics at the Parker Institute for Cancer Immunotherapy.
Over the years, Alfa said he became interested in the problem of complex biomarkers in tumor biology. Cancer is not driven by a single genetic driver, rather, it stems from multiple factors. Alfa followed the research developments in cancer immunotherapy. What he saw was a disconnect between exciting preclinical data and disappointing results when those therapies reach human testing. He also saw a path to a solution through machine learning.
While Recursion has cancer drug research, Alfa said Noetick is unrelated to anything at his former company. The cancer data that the startup generates and analyzes are different than what Recursion is working with. The tools are different too. Noetik was founded in mid-2022. The startup is not the kind of company that could have come together as recently as a few years ago, Alfa said. In that span of time, the field has made advances in spatial technologies, software that gathers information from pathology specimens, and machine learning. In less than one year, Noetik says it has generated hundreds of terabytes of human data to support its research.
Alfa declined to discuss details about Noetik’s research, but speaking generally he said the technology enables Noetik to see spatial interactions of immune cells and tumors. From these interactions, Alfa said the company is able to see drug targets. While Alfa isn’t disclosing any targets, he said the platform can address different types of tumor biology with multiple drug modalities. The company aims to develop precision immunotherapies.
Noetik’s seed financing was led by DCVC. Other participants in the round include Zetta Venture Partners, 11.2 Capital, Catalio Capital Management, Epic Ventures, Intermountain Ventures, North South Ventures, CJNV BioVentures, Viswa Colluru and Hummingbird Nomads Fund, and Recursion CEO Chris Gibson and Michael Secora, the company’s chief financial officer. Alfa said Noetik will use the capital to expand the team and bring in enough tumor samples for the platform to demonstrate proof of concept.
“Proof of concept is showing that we can train models to understand fundamental tumor biology from these samples and start learning new biological insights,” he said. “I can’t share much. But we’re already seeing we can train models on this information to the point that they’re learning about the biology.”
Photo by Noetik
