Send us a text

A new collaboration between Yale and Google just released C2S-Scale, a 27-billion parameter foundation model that's learned to "read" cells by converting gene expression into text.

It's already predicted a validated cancer therapy combination that worked in the lab. But here's the question that matters to our community: What can this tool actually do for rare diseases? 💡

In this episode of March Forward, we're not just celebrating the hype; we're asking the hard questions that patients and families need answered:

🔬 Can it work reliably for rare diseases when it was trained mostly on common conditions? We explore how the model handles the "rare data problem" and whether its predictions will be relevant or misleading for out-of-distribution cases.

💭 Does it understand genetic mutations, or just their effects? We break down what it means that the model reads RNA (gene expression) rather than DNA, and what that limitation means for understanding the root cause of disease.

🎯 When AI suggests a therapy, is it finding real biology or just statistical patterns? We analyze the celebrated silmitasertib discovery to understand what "AI-driven therapeutic discovery" actually means and whether we can trust the mechanisms it identifies.

✨ What's missing from the picture? The model currently only works with gene expression data. We discuss the roadmap (and challenges) for integrating protein, metabolic, and other biological layers for a complete understanding.

💪 Can it help define "healthy" when we don't have a baseline? For many rare conditions, we don't even know what a healthy cell should look like. We explore the revolutionary potential of using this AI to generate reference points.

Today, we're diving deep into the architecture, the limitations, and the unanswered questions, because understanding what this technology can and cannot do is crucial for making informed decisions about rare disease research.

Don't forget to subscribe to March Forward for more insights from our community run by patients, for patients. Share your thoughts or questions in the comments below; we're all in this together. 💙

Visit march.ai to connect with others, share experiences, and stay informed.

#RareDisease #AIinMedicine #C2SScale #SingleCellBiology #GeneticDisorders #PatientCommunity #MedicalResearch #MarchForward #MarchCommunity #RareDiseaseAwareness #HealthPodcast #PrecisionMedicine #ComputationalBiology #TherapeuticDiscovery #PatientAdvocacy #MedicalAI #Genomics #DrugDiscovery

March Forward - Podcast by March.ai