Tahoe Therapeutics has raised $30 million to embark on what it calls a transformative leap in precision medicine: the creation of the largest foundational dataset for training AI-powered Virtual Cell Models. The initiative will generate an unprecedented one billion single-cell datapoints, mapping one million drug-patient interactions — a scale never before achieved — with the goal of uncovering new precision cancer therapies and beyond. In a bold strategic move, Tahoe plans to share this dataset exclusively with a single partner, a pharmaceutical or AI firm equipped to help translate the data into real-world clinical breakthroughs.
The funding round was led by Amplify Partners, joined by a prominent lineup of backers that includes Databricks Ventures, Wing Venture Capital, General Catalyst, Civilization Ventures, Conviction, Mubadala Capital Ventures, and AIX Ventures. This capital infusion builds on Tahoe’s growing influence in the field, coming just months after the release of Tahoe-100M, the world’s first gigascale perturbative single-cell dataset. That earlier dataset, open-sourced and downloaded nearly 100,000 times, has already catalyzed the discovery of promising therapeutic candidates for major cancer subtypes and revealed new drug targets across modalities. It has become a critical resource for AI labs and research institutions working on virtual cell models.
Tahoe’s next phase expands its ambition tenfold. By mapping the interactions of tens of thousands of drug molecules with human biology at single-cell resolution, the company seeks to extend the frontier of biological foundation models, aiming to reduce the historically high failure rates of clinical trials. CEO Nima Alidoust frames the effort as the biological equivalent of AI’s “GPT moment,” in which massive datasets catalyze leaps in model performance. This advance, he says, will combine deep biological insight with computational power to deliver new medicines more efficiently and with far greater likelihood of success.
The company’s approach integrates single-cell genomics, advanced machine learning, and scalable drug screening across diverse patient samples. Building on scientific breakthroughs at UCSF, Tahoe’s platform generates the raw, high-dimensional biological data required to train disease-relevant AI models of human cells. The new funding will not only drive the dataset’s creation but also advance Tahoe’s own therapeutic programs toward clinical trials. By aligning with one strategic partner, Tahoe hopes to fuse its massive biological datasets with complementary clinical or AI modeling capabilities to develop the first medicines born from virtual cell models.
Amplify Partners’ Sunil Dhaliwal highlights the urgency and potential of Tahoe’s work, noting that while structural models have made strides in molecular design, they have not reliably translated to clinical wins. By shifting the focus to high-resolution, cell-based datasets that reflect real patient biology, Tahoe is positioning itself to break through a longstanding bottleneck in drug development. If successful, this effort could mark a turning point in precision medicine — one where AI models of human biology are no longer just predictive tools, but active engines of therapeutic discovery and clinical success.
Leave a Reply