Good to see folks going back to the basics. Been saying this for a while, that the whole AI virtual cell space has been putting parameters first & expt design second (or not at all) when the latter is the most important to match your questions of interest. 1/

Unlike “virtual cell models," Rhaister goes back to the basics: it builds a minimalist perturbation model from the ground up, directly on summary statistics of the data.

@tahoe_ai team doing what they do best! A key takeaway from VCC’25 for us was that in the current data regimes, elegant statistical models match if not surpass far more complex transformer-based architectures! “Going back to the basics” is such a great title for this… to say nothing of the 2M cell dataset! Kudos…

These results demonstrate we can accomplish a lot by going back to basics and building models that, by design, reflect the statistics of the underlying data. Rhaister shows that scaling the right data is far more valuable than scaling parameters.

A more in-depth tweetorial from @vallens

Also will once again iterate that one can develop simple approaches that can predict perturbations that are in distribution very well without ever learning a causal model of gene regulation. 2/

It has been a wonderful surprise for us to see such an interpretable, inexpensive model (trained in seconds, predictions in milliseconds), accomplish what virtual cell models (typically with far more complex architectures) promised to eventually do.

And it is capable of predicting more complex drug phenotypes such as sensitivity in cancer cells far beyond simple baselines. And there is room to make it much better, so stay tuned.

With a handful of example perturbations in a new context, Rhaister predicts responses for other perturbations with accuracies within experimental noise, exceeding state of the art virtual cell models in performance.

@nalidoust So excited for this to be finally out! Incredible work led by @vallens . Rhaister allowed us to do launch hundreds of dataset scaling experiments in hours in what would have taken months with the previous generation of perturbation models 🚀🚀

It is the first model we have seen that performs significantly beyond mean baselines in a zero-shot setting; a task commonly proposed as a promise of virtual cell models.

Despite its simplicity, it is the first model that shows consistent scaling with more perturbation data.

Read more Paper: https://tahoebio-assets.com/rhaister-manuscript.pdf Model and datasets: https://huggingface.co/collections/tahoebio/rhaister

What excites us more is what comes next: fast and interpretable, Rhaister is far better suited to advance biological reasoning in close iteration with the emerging agentic workflows.

We show that by testing it against our very unique Emerald Bay dataset, generated using our Mosaic platform, measuring 5-day sensitivity of cancer models to various drugs. And we are open-sourcing that dataset along with Rhaister.

@nalidoust Here's to another moon landing 🚀💙

@nalidoust Huge advancement for the field 🔥💯

@nalidoust Let’s goooo! 🚀🚀

@nalidoust 🔥🔥🚀

@nalidoust 🚀🚀🚀