How AI is Driving Drug Discovery: Xaira Therapeutics' Marc Tessier Lavigne

Mark Tessier Levine (Zara Therapeutics CEO) identifies the core drivers of institutional success as a combination of clarity of purpose, culture, and leadership.

• Levine cites Genentech (acquired by Roche for $47B in 2009), Rockefeller University, and Stanford University as examples where high standards for personnel define excellence.
• At Rockefeller, the hiring bar is set so high that the institution made zero faculty offers in one year despite receiving 1,000 applicants.
• The organizational culture at these entities prioritizes recruiting individuals capable of fundamental change rather than incremental scientific gains.

Levine's personal trajectory shifted from pure neuroscience to drug development through three converging events:

• A funding relationship with the Paralyzed Veterans of America spinal cord research foundation highlighted the potential for applied science.
• The paralysis of his father from a stroke made the research personally urgent.
• A recruitment to Genentech provided executive training under CEO Art Levinson on how to run teams and build organizations.

The drug discovery industry faces significant inefficiencies with a clinical success rate that has remained stagnant at 10% for the past 20 years.

• Current attrition rates are high, with approximately nine out of ten drugs failing after entering the clinic.
• The current process relies heavily on intuition and empiricism, costing between $2 billion and $4 billion per successful drug.
• The industry is currently in an "artisanal" stage regarding target selection, drug design, and patient selection.

Zara Therapeutics aims to transform drug discovery by replacing trial-and-error with AI-driven in silico design.

• The company's vision is to "design the needle" rather than searching for it in a haystack of biological data.
• Zara is executing a staged approach across three specific domains:
  • Leveraging David Baker's AI models to design antibody drugs, specifically targeting "undruggable" proteins.
  • Developing biology foundation models to understand complex biological systems.
  • Creating patient representation models to improve patient stratification and clinical trial success.
• The company is combining frontier AI with high-throughput biological data generation.
• Zara has raised $1 billion to fund a massive build-out of infrastructure integrating these three components.

Zara differentiates itself by addressing all three pillars of drug discovery simultaneously, whereas many competitors focus on narrow segments.

• This approach combines frontier AI, massive-scale biology, and deep drug discovery experience.
• The company is actively moving forward with a therapeutic pipeline despite the rapid evolution of underlying AI models.
• Leadership philosophy emphasizes not waiting for perfect technology, adhering to the principle that "the best is the enemy of the good."
• Industry accelerations are already visible, with some companies reducing timelines for drug development from a 5-year average to under 2 years.