Insitro Adds $143M to Fuel Machine Learning-Driven Drug Development

Xconomy San Francisco — 

A stem cell biologist, a gene editing expert, and a data engineer walk into a room…and that’s just a few of the people Daphne Koller has recruited in the past two years to Insitro. The South San Francisco-based startup has grand plans to reevaluate and, potentially, reinvent the drug discovery and development process using the latest tools in the cell biology, bioengineering, and data science armamentariums.

On Tuesday the company announced a syndicate of investors added $143 million to its coffers in a Series B financing led by Andreessen Horowitz (a16z). Those funds more than double the total the company previously raised to fuel its ambitious plans: Its Series A financing was upwards of $100 million.

Want more AI, machine learning, and drug discovery content? Network and engage with industry leaders online this June. Learn more.

Koller (pictured), its CEO and founder, was an “n of 1” back in May 2018, when the company launched. Since then, bringing together the right group of people with the various areas of expertise needed to create a platform that combines leading-edge technologies at production scale has been a “major focus” for the former Stanford University professor, longtime machine learning researcher, and entrepreneur.

“They each bring their own unique perspective into what we’re doing because of the breadth of the goal that we have,” she said in an interview. “What I’ve found is that they don’t just come up with better solutions; they do, in fact, come up with problems that are not the ones you would have thought about, and it’s really been an incredible well of creativity for what the company does.”

The Insitro team now totals about 65 people, including Ajamete Kaykas, who formerly led an early target discovery team focused on neuroscience at Novartis (NYSE: NVS), as chief technology officer; Mary Rozenman, previously senior vice president of corporate development and strategy at Aimmune Therapeutics (NASDAQ: AIMT), as chief business and financial officer; and Serafim Batzoglou, most recently vice president of applied and computational biology at Illumina (NASDAQ: ILMN), as chief data officer.

By year’s end Koller says more hires are slated to bring the Insitro workforce to about 95. (The company’s name is a mashup of in vitro and in silico, referencing, respectively, the biological experiments done in the lab and those done on computers.)

Today the company, in response to the engineering challenge it has set for itself of creating the right amount and quality of biological data to feed robust predictive disease models at scale, is finalizing what Koller describes as version 1.0 of its technology platform.

“We’re starting to look at how we can take even the capabilities that we have and build some of those disease model systems for a certain category of diseases so that we can start to identify targets that we believe in,” she said. “We already have some of those, but it’s very early stage, and we don’t yet have the confidence or the conviction to start turning those into drugs, [but] that’s something we’ll be considering over the next few months.”

Koller says she is being careful to focus on therapeutic areas where—for the time being, at least—the tools the company is creating can be most useful.

“We’re very cognizant of the fact that this isn’t a silver bullet that will work for every single disease,” she said. “The ones that our particular approach that we’ve employed so far has been, we think, especially useful for are diseases that have a strong genetic basis, so you can see some level of how the disease burden that comes in from the genetics can align, potentially, with what we see in terms of the outcomes for patients, and where the cellular systems are tractable, because not every cell type can be differentiated from [induced pluripotent stem] cells; and, furthermore, some diseases are so complex and systemic that it might require 30 different cell types working together all in exactly the right way to get something that looks like the disease, and we’re nowhere close to being able to engineer something like that.”

Insitro has focused its initial energies on liver diseases, prompted by a partnership it struck about a year ago with Gilead Sciences (NASDAQ: GILD), and disorders of the central nervous system.

These starting points aren’t exactly low hanging fruit. Just this year companies working on late-stage treatments for the liver condition Gilead and Insitro have teamed up to tackle, nonalcoholic steatohepatitis (NASH), have reported failures and delays in their efforts to develop an FDA-approved therapy.

And developing treatments that work in the field of neuroscience has proven perhaps even more challenging, with scientists struggling to understand the relationship between disease and pathophysiology as well as to conduct clinical trials that reach statistical significance given the variable and subjective nature of the endpoints established.

Koller says Insitro’s cellular models may help ameliorate some of those obstacles.

“What is a depressed mouse? What is a mouse with ADHD?” she said. “This is an area where animal models don’t work very well … There’s a real opportunity to create a different and, hopefully, more translatable disease model for some of these indications.”

In keeping with its strategy since its launch, Koller says she plans to build Insitro into a company that develops internal drug candidates, rather than one that solely works with biopharmas to advance their programs.

That’s a tall and expensive order, and she says it was essential to find financial backers with a long-term view of what the firm is trying to accomplish.

“There are many investors, especially in this kind of growth phase, who are looking for a quick exit: you make an investment, 12 to 18 months later the company goes public, they achieve liquidity, they sell the stock and they’re happy,” she said. “This is exactly not the kind of investor that we were looking for, and we’re happy to be able to attract a group that is looking for something completely different. They’re looking for investments that might take a longer amount of time, but, hopefully, for a very high reward, because they’re looking to build transformative companies.”

Insitro added some new investors as part of its Series B round, including the Canada Pension Plan Investment Board (CPP Investments), funds and accounts advised by T. Rowe Price Associates, funds managed by BlackRock, Casdin Capital, HOF Capital, WuXi AppTec’s Corporate Venture Fund, and other undisclosed investors. Earlier investors ARCH Venture Partners, Foresite Capital, GV (formerly Google Ventures), Third Rock Ventures, Two Sigma Ventures, and Alexandria Venture Investments also participated in the financing.

As part of the deal a16z’s Vijay Pande, a general partner and founding investor of the firm’s Bio Fund, joins Insitro’s board of directors.