Why Most Drugs Fail, and Why Virtual Tissue Models Beat Virtual Cell Models

Naren Tallapragada (co-founder of Tessel) joins to talk about the industry of modeling biology. While the field races to build virtual cell models, his team builds virtual tissue models, betting that predicting what an organ actually does matters more than predicting what genes a cell expresses. The goal: not just whether a drug works, but who it works for.In this episode, we discuss:Why 90% of drugs fail in clinical trials, and why neuro and Alzheimer's failure rates are even worseThe multi-scale modeling problem: molecules, genes, cells, tissues, organs, and why picking the right level of resolution matters more than picking the "best" oneThe electron-and-lightbulb argument for why virtual cell models may be a beautiful solution to the wrong problemOrganoids vs. animal models: where human cell models beat animals, and where both fail (like modeling behavior in mental health)Why Naren thinks defensibility in AI-driven biology comes from proprietary multi-organ data, not model architectureThe case for precision medicine, and why it's been such a brutal business model to actually execute onCredits:Created by Greg Kubin and Matias SerebrinskyHost: Matias and GregProduced by Nico V. ReyFind us at businesstrip.fm and psymed.venturesFollow us on Instagram and Twitter!Theme music by Dorian LoveAdditional Music: Distant Daze by Zack Frank