
A new AI startup, launched with a trio of Fred Hutch vets, wants to give a 'zoomed-in' view of cancer cells
Ali Ansary, a trained physician, had realized by 2018 that cancers would no longer be treated just with traditional chemotherapy and radiations. Immunotherapy was becoming big, and yet there was no way for doctors to get a complete picture of how immune cells worked at the cellular level when used for cancer therapy, he says.
“How do we get to know more information about this system that is more than inflammatory markers?” he wondered.
Then, through a Seattle venture capital community, he met three Fred Hutch researchers in Greg Finak, Raphael Gottardo and Evan Green. After several conversations about the problems they saw in the space, Ozette Technologies was born the next year.
This Seattle-based startup wants to automate analysis of data collected from a single cell and do away with the traditional way of doing it manually, joining the cadre of AI biotechs jumping into the space recently. It received a $26 million Series A funding, the company announced Thursday.
“The (cellular) data is so complex that we miss a lot of information when done manually,” says Ansary, Ozette’s CEO. “In fact, 90% of the data is sitting on the table and a lot of people say we do not even know how to look at this kind of data. It is so complex.”
Results from cellular experiments are typically analyzed manually, he says, with a process looking something like computational biologists going over the data before sending it to data scientists, who would in turn send it to statisticians and so forth. But Ozette claims it has devised a machine learning-based platform that completes months of analysis in a matter of days.
“We are using machine learning to uncover all the different parts of the data that comes out of experiments,” he said.
The company saw one of its initial successes when it analyzed the tumor cells of skin cancer patients taking PD-1 inhibitors, Ansary said. The Ozette research team was able to give a zoomed-in picture of the tumor environment and published their findings in the journal Patterns. The research showed that some aspects of the tumors were replicated in the bloodstream, which led to the discovery of novel biomarkers specific to patients who respond to the treatment, Ansary said.
“Imagine all the people in the world who have a tumor,” Ansary said. “You cannot go and always do biopsies on the tumor. In that case, you can extract valuable information around what the immune system is doing in that tumor, by just looking at the blood.”
The funding was led by Madrona Ventures with other participating investors such as Cercano Management (formerly Vulcan Capital), M12, Microsoft’s venture fund, Alexandria Venture Investments, OCV Partners and Duke University. The new funding will go towards expanding the lab space that the company recently launched and hiring more data scientists and web lab scientists.
“The market is still at an early stage, because it is only in that last handful of years, we have been able to get down to the single-cell level and distinguish a particular cell from another cell,” said Matt McIlwain, managing director at Madrona Ventures, which had also poured in $6 million in seed funding in 2020.
“Ultimately,” Ansary says, “whether someone is healthy or sick, we want Ozette to be a key part in giving insights into what’s happening.”