Machine learning, data science, and other computational techniques are being used by a growing number of bioform startups to find new drug targets, develop new therapeutics, and support their testing.
Venture capitalists – including those who have not touched science before – are paying attention.
As more and more biotech companies move into once-owned technologies, traditional venture technology firms are also turning to life sciences, including Silicon Valley heavyweight Andreessen Horowitz (a16z) and Madrona Venture Group of Seattle.
New technologies, such as artificial intelligence, are changing the way drugs are discovered and developed, and more and more biotech startups have adopted engineering thinking. Investors who in the past may have considered life sciences too risky are now experiencing opportunities.
Madrona’s first investments in what she calls a “crossroads of innovation” began through her technical connections.
“Common are the people we’ve worked with before who are also passionate about life sciences and healthcare,” said Matt McIlwayne, Madrona’s managing director.
Suyal Patel, for example, moved into life sciences after founding Isilon Systems with the support of Madrona, which sold EMC for $ 2.25 billion more than a decade ago. In 2016, he co-founded Nautilus Biotechnology for protein analysis, and Madrona moved in with him, investing in a startup before it went public last year as a result of a $ 345 million SPAC merger.
Patel is one of the growing cadres of biotech and life sciences executives who have made the leap from software startups.
For example: Terry Myerson, a former senior executive at Microsoft, now heads Truveta, a Seattle-based startup that seeks to consolidate data from medical records.
Insitro founder Daphne Kohler was a co-founder of Coursera and a professor of computer science at Stanford. She now heads Insitro, a giant startup for drug discovery and development supported by a16z.
“I see more and more people moving from technology to such companies,” Pande said. “They see opportunities, and, frankly, they are driven by a mission, a desire to have a huge impact on human health.”
McIlwine also has a personal passion for the life sciences. Both of his parents survived cancer, and he is a board member and previous chairman of the Fred Hutchinson Cancer Research Center.
Madrona has strengthened her life sciences team and her connections with scientists from Seattle. Companies supported by Madrona include A-Alpha Bio, derived from the Institute of Protein Design at the University of Washington, Ozette, Fred Hutch, and Modulus Therapeutics, a field of cell therapy at the Allen Institute of Artificial Intelligence.
A sense of opportunity
Mark Andreessen, co-founder of Netscape, who helped launch the a16z, has historically not been a fan of investing in life sciences. Andreessen said that the firm “will not be engaged in biology, will not be engaged in life sciences, will not be engaged in health care,” – said Vijay Pande, a former professor of biophysics at Stanford, who oversees the funds a16’z Bio.
“Many life sciences have been dominated by binary risk, where something will work or not. It’s almost like a science lottery. And healthcare was dominated by services – such low incomes, ”Pande said. “Both of these things really run counter to the technological spirit.”
But in 2015, the venture company attracted Pande and launched its first $ 200 million biofund. More startups in the life sciences have adopted technological and engineering thinking, offering the opportunity for faster returns. The emergence of more platform startups contracting with major pharmaceutical companies has also been attractive, Pande said.
In January this year, the company raised $ 1.5 billion for its fourth biofund. The firm has invested in many life science companies such as Insitro and Nautilus, as well as in healthcare technology companies, where AI has the potential to reduce the cost of providing services.
Investors are also paying close attention to the Seattle area with its strong life science institutes and powerful technology companies as the main region for startups at the merger site. “Seattle has placed both jaws in a way that not many other places do,” Pande said.
Biopharma companies have traditionally faced a slow pace of clinical trials and a pull to the end. Many treatments just don’t allow it.
“This world seemed kind of scary to us, to be honest,” McLaughlin said.
Madrona began to notice the emergence of more platform-oriented companies with new ways of collecting and analyzing biological data. “It felt more in our wheelhouse,” McIlwayne said.
Drug discovery is a “barbell problem,” Pande said. One side of the bar is research in the early stages and the other is clinical trials. Both sides are dear.
“The middle part, in fact, is where traditional pharmaceuticals have worked pretty well. Given the goal, they can come up with something [a testable drug] pretty soon, “Pande said. New technologies such as AI can help in the middle. But where it really has the greatest potential, it’s a big barbell point in the beginning, he said.
By identifying the best targets for drugs, AI has the potential to attract the best therapeutic candidates to the clinic and eliminate the waste of failed clinical trials, Pande said.
Human biology is increasingly understood as a collection of rich data sets of genes, proteins, cells and their interactions.
With a deeper understanding, “this transition from science to technology, from discovery to design, can happen,” Pande said. “I think that’s the aspect that really excites people.”
Pande added: “People are doing drugs all the time. The question is, are they changing the way drugs are produced? ”
Since 2014, investors have invested $ 5.1 billion in 103 venture rounds for biotech startups with AI and machine learning – and $ 3.3 billion came in just last year, according to Endpoints.
More than a new technical step
AI is not just the next step in the march of biotechnology, similar to stages such as gene editing and next-generation sequencing, Pande said. AI contributes to a greater shift in thinking from discovery to design.
“It’s not just the availability of additional tools in the craft process,” Pande said. “It actually changes the nature of the process. It’s actually industrialization, which makes it more like a factory than a craft business. ”
As pharmaceutical companies focus on synthetic biology and drug design based on artificial intelligence, they are increasingly similar to engineering firms, he said.
“This shift is a huge cultural clash,” he added.
Bringing together biologists and technical experts for effective collaboration is a challenge, GeekWire Color previously told Insitro.
Meanwhile, more biologists are training with double experience. “I think they will really make the biggest contribution,” he said.
Development of biotechnology
Life science companies are working more cautiously than technology companies, in part because of their more limited regulatory environment and the need for a stronger ethical focus, McIlwayne said. They also need to pay more attention to intellectual property. But there is room for change.
“There are certain circumstances when you can bring more flexible, iterative, continuous thinking that lives in the world of software and technology,” McIlwayne said.
Pande sees that some of these changes begin to occur when scientists better understand biology.
“In technology or any other engineering discipline, if something doesn’t work, you can replicate and improve. And this is actually what we are starting to see in technology that is also used in drug development, ”Pande said.
Companies in technology, health and life sciences are increasingly finding a match. They are “moving to the center,” Pande said. According to a recent CB Insights report, digital healthcare startups raised a record $ 57.2 billion last year, up 79% from 2020.
Even the early days
Companies working on the convergence of life sciences and technology are young, and their first treatments have only recently undergone clinical trials. Their approaches may seem opaque and secretive to outside scholars. Are the promises of discoveries based on artificial intelligence inflated?
“I think it’s too early,” Pande said. “Think of the 20-year arc of the Internet, maybe from 1995 to 2015 or so. There were the first days when there were skeptics who said, “It’s funny, you’ll never buy dog food online, I can give you 10 reasons why it’s impossible.”
According to Pande, this arc now lasts about five years, at the stage of many companies before the field consolidates and dominant players appear.
McLean holds a similar view. While some of Madrona’s early investments in another area, AR / VR, have declined, the VR Rec Room startup reached an estimate of $ 3.5 billion last December.
“It’s so important for us to have a prepared mind, to be attentive to new categories like this intersection of an innovation theme,” McIlwayne said. “The first day for the long term” is our strategy, so we will be too early. That’s who we are. That’s what we do. ” Sometimes that means a few companies won’t succeed, McElwain said, but others will.