A Billion-Dollar Bet on AI-First Drug Development

I'm Daniel Levine, and this is the Bio Report. Despite the emergence of new modalities and drug development technologies, the cost and time to produce new therapies has changed little, and failure rates remain high. Zara aims to change that with a systemic AI driven approach that tackles three pervasive bottlenecks choosing the right targets, designing the right molecules, and matching the right patients by running as much work as possible in silico and using high dimensional causal datasets to train virtual cell foundation models. The company is initially focusing on high value, historically undruggable targets and ultimately on building a pipeline of differentiated biologics. We spoke to Marc Tessier Lavigne, co founder and CEO of Xara, about applying end to end AI across target discovery, molecular design, and patient stratification, the company's more than $1,000,000,000 in funding, and how it seeks to enable a new generation of scientists fluent in both AI and biology. Mark, thanks for joining us. Thank you, Danny. It's great to be here. We're gonna talk about Xara, its vision of itself as a new generation drug company, and and its efforts to remake pharmaceutical r and d with an end to end AI approach. Let's start with the problem you're seeking to address. What's wrong with the traditional approach to drug discovery and development? Well, we made great strides, over the past several decades with, drug discovery and development, especially in opening up new fields, new modalities. You know, think about, bio conjugates, armed antibodies. Think about, you know, RNA vaccines, and the like. But at the same time, the overall productivity of the sector, has not, improved much. If you go back twenty years, the figures that, you know, were discussed was that on average from starting to work on a target to making a drug and then pushing it through clinical trials to FDA approval would take on average about thirteen years. Attrition in the clinic, you know, drug centering the clinic ninety percent to ninety five percent would fail. And, and the cost was, in the billions of dollars. Dial forward twenty years and the numbers pretty much haven't budged. So although we there have been great, successes over the past several decades, especially in opening up new fields, at the same time, as a whole, we know first of all, there are opportunities that we're we're not, able to access. And secondly, we just have to do a better job of of fixing, the overall process, reducing the timelines, increasing success rates, reducing attrition, and, and thereby lowering costs. The way we do business today just isn't sustainable, in the long run. And so, there's a need to rethink this whole process of drug discovery and development. What's the potential for an integrated AI approach to fix that? And is it fixing it by addressing the high failure rate or there are other ways that it's going to reduce cost and compress timelines? I think you you have to use …