Andrew Radin orchestrates his AI discovery platform's pivot to R&D with a harmonious name change and eyes on the clinic
Andrew Radin was a straight A student. So when he stopped turning in assignments in Nigam Shah’s class at Stanford University, the professor knew something was up.
“If you don’t hand in assignments, you’re gonna fail,” Radin recalls Shah saying one day after class. “Everything alright at home?”
Vijay PandeLittle did Shah know that a project Radin did for another class had piqued the interest of Andreessen Horowitz’s Vijay Pande, who also taught at the university. As a biomedical informatics student, Radin was looking at how technology could be used to comb through large swaths of data and find drug-target matches — and Pande offered him the first investment from his new $200 million biotech fund to turn the platform into a company.
So Radin dropped his studies, and in 2015 officially launched AI-focused twoXAR. Shah, who’s now on the scientific advisory board, called it the “best F ever.”
The company now boasts multiple biotech partnerships and nearly as many programs as it has employees (19 staffers versus 18 programs) — all accomplished with a modest $14 million in VC funds. And as it transitions from partnering on drug discovery research to advancing its own pipeline, it’s changing its name to Aria Pharmaceuticals.
“An aria, it’s this capstone of the musical piece that’s typically reserved for a very talented performer,” Radin, the firm’s CEO, told Endpoints News. “And yet, the aria doesn’t stand alone. It works within the context of a larger piece of music.”
The “aria” in this case, is the company’s drug discovery platform, he said: “exceptionally gifted in its own right,” but more powerful when coupled with expertise and the right team to bring drugs to patients.
To date, Aria has identified 18 potential candidates for complex diseases like lupus, glioblastoma, chronic kidney disease and glaucoma, four of which are being developed with partners. No candidates have reached the clinic, but the team is preparing for INDs in several programs.
Using AI technology, Aria has been able to complete predictions, select hits and begin in vivo testing in an average of four weeks — a process which normally takes years using traditional discovery methods, according to Radin.
The machine learning space is packed with players looking to revolutionize the way new drugs are found and developed. Just last week, Exscientia raised a fresh half-billion dollars for its AI platform and pipeline. Microsoft backed a new AI startup called 1910 Genetics back in March. And earlier that month, Daphne Koller pulled in $400 million for her machine learning startup, insitro.
“What’s special about our methodologies, is … people will typically use one type of data or maybe, you know, a few types of data together in concert to try to make a discovery,” Radin said. “We incorporate dozens of unrelated datasets in our prediction systems, and what that does is give us a very wide swath of information to basically uncover new models of pathogenesis.”
To lead the charge, Aria recently tapped Mark Eller, former head of R&D at Jazz Pharmaceuticals, as senior VP of R&D; and Anjali Pandey, who was most recently CSO at BridgeBio subsidiaries TheRas and Ferro, as senior VP of nonclinical R&D and chemistry.