AI-led Drug Development a Reality Now

AI-led Drug Development a Reality Now

Right in the midst of the raging debate on how artificial intelligence (AI) would impact pharma, medicine and healthcare management, a new startup has set its sight on using GenAI-led computational methods to revolutionize drug discovery itself. At the forefront is Xaira Therapeutics, which emerged from stealth mode, with a billion dollars in its bank. 

At the helm of the new venture is a former Stanford president and chief scientific officer at Genentech Marc Tessier-Lavigne, who believes that his company would develop drugs that were once impossible to make minus the revolutionary breakthroughs in AI. By the looks of it a bevy of investors are also convinced of the idea. 

Big tech gets big funding from VCs

Besides ARCH Venture Partners and Foresite Labs (a part of Foresite Capital) who incubated Xaira Therapeutics, others who’ve joined the funding includes Sequoia Capital, Lightspeed Venture Partners, Menlo Ventures, Two Sigma Ventures, F-Prime, NEA, Lux Capital, and SV Angel. Total funding is around $1 billion and counting. 

According to Tessier-Lavigne, the company did a huge capital raise, after remaining in stealth mode for over six months, as they believe GenAI and associated tech was at an inflection point where it could transform biotechnology. The initial forays into foundational models came from the University of Washington’s Institute of Protein Design helmed by David Baker. 

The right ideas and the right folks

That Baker is one of Xaira’s co-founders adds to the euphoria. Baker’s models are aligned to the diffusion models that function behind the scenes for image generates such as DALL-E, though instead of creating arm, these are aimed at designing molecular structures that can be created in a three-dimensional physical world. 

Of course, none of those involved are under any illusion that work on these GenAI models and its impact on drug discovery could be a long-drawn process. All that they’re convinced of is that Baker’s models can revolutionize data design though how GenAI could potentially change biotech is still being researched. 

The challenges before AI-led biotech

One of the big challenges around modelling is the availability of quality data. While training GenAI in other segments of industry involves data from consumers, biotech doesn’t allow this to any great extent. Therefore, one could say that the first task of Xaira could be to create datasets that could then drive model development. 

Of course, Xaira isn’t the first and won’t be the last to attempt such an innovation. Recursion is another company that’s seeking GenAI help to design drugs. This publicly listed company as well as Genesis Therapeutics are already in the field, with the latter having raised a Series-B funding of $200 million. 

Neither of these companies have a definite date for having a drug on the market and the case is no different for Xaira. However, in their case, the bunch of investors who pumped in all that money are ready to play the waiting game. They are aware that drug discovery is a costly affair and becomes doubly so when it is AI-led. 

Analysts are skeptical, but VCs aren’t

Of course, market analysts are skeptical, not because AI-led drug discovery is still a distant dream, though not a piped one. Their challenge is more immediate and relates to the appointment of Tessier-Lavigne as CEO, given that his laboratory at Genentech made the news for the wrong reasons, one of which was to do with manipulating research data. 

Though he was cleared of any wrongdoing, Tessier-Lavigne quit Stanford due to the “distractions” caused by the investigations and noted that he was doing so “for the good of the University”. 

Of course, Xaira investors aren’t too concerned with the past and are hoping that with a rich purse, they would be able to crack the code and get the first AI-led drug into the markets sooner than later. 


Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *