Wading deeper into target discovery and validation, Immunai bags $60M to beef up both AI algorithms and lab benches
Noam Solomon and Luis Voloch set out, when they launched Immunai in 2018, to map out the immune system cell by cell.
The two engineers — who had met at MIT and like to illustrate their company’s stature by pointing out the number of employees who had been trained at Palantir, Google or Facebook — saw the potential in collating all the information churned out by a single cell-sequencing multi-omics platform. You can tell drugmakers what exactly is in their cell therapy products, profile for academics the immune response to their experimental treatments, or even suggest new biomarkers that seem to be relevant in a disease.
“Every sample that we sequence is more than a terabyte of information,” Solomon, the CEO, told Endpoints News. “So if we are talking about the database on the order of thousands and even tens of thousands of samples, we’re talking about massive databases, and running computations on such a large database requires differentiated capabilities.”
But as Immunai signed more deals, they realized that something else was missing. Partners often didn’t have a “data-driven” way to validate the targets, Solomon said, and they believe they could help by offering one more step beyond mining insights from their big database.
So having raised $20 million in seed funding last May, the duo have bagged a $60 million Series A to move into what they call functional genomics.
“It moves us from the correlative analysis — which can still be very powerful — into things that are like causal validation,” CTO Voloch said. “We’re building our whole infrastructure to be able to do this at a scale and sophistication that no one has right now.”
The setup will be complete with a wet lab equipped to conduct both in vivo and in vitro testing. After its algorithms identify a genetic signature potentially tied to tumor response, for example, scientists at the company could then design experiments to tease out the target’s value.
That’s the promise, at least, for the collaborators at Baylor College of Medicine and the pharma customers who gained access to the AI platform through a deal with 10x Genomics.
Immunai has so far accrued samples “in the thousands” with a 70-strong team spread across 15 countries, and there are plans to double or even triple the group as they push to sequence more.
But Voloch emphasized that it’s the ability to look at individual cells — sometimes longitudinally, tracking T cell clonal expansion over the course of treatment — that matters most.
“We have a database of cells so people can query for cells of specific features,” he said. “And that’s how we’ve been thinking about it and so all of our machine learning and AI is largely at that cell unit.”
Schusterman Family Investments, Duquesne Family Office, Catalio Capital Management and Dexcel Pharma led the round, while existing investors Viola Ventures and TLV Partners participated.