In a recent 26-minute episode of the RARECast podcast, our co-founder and CEO, Jean-Philippe Vert, discusses how we are using AI to fundamentally rewrite the rules of research and development for rare diseases.

Historically, drug development has been a costly, trial-and-error effort where nine out of ten clinical programs fail despite major scientific advances. A major reason for this massive failure rate is that biological information is often fragmented in silos, with traditional R&D relying heavily on narrow, task-specific datasets. This challenge becomes especially severe when targeting rare diseases, which suffer from a lack of patients, scarce data, poor preclinical models, and extremely difficult development economics.

To solve this, we at Bioptimus are building a foundation model that breaks down these silos. By integrating multimodal and multiscale biological data into a single, comprehensive body of knowledge, our AI is designed to make sense of the inherent messiness of biology.

In this episode, Jean-Philippe explores:

• Overcoming Data Scarcity: How our foundation models offer immense promise for rare diseases by operating effectively even when patient numbers and traditional datasets are highly limited.
• The Messiness of Biology: Why traditional R&D struggles to capture the immense complexity of human biology, and how our holistic AI approach bridges that gap.
• Repurposing Existing Drugs: How uncovering hidden biological similarities between different conditions could allow researchers to quickly and safely repurpose existing drugs for new rare disease treatments.

Listen to the full episode to learn how we are leveraging AI to bring hope and accelerated treatments to the rare disease community.

‍