Collaborating with NVIDIA on biological foundation models

A month ago we, at Tahoe (formerly Vevo) announced that we have generated the Tahoe-100M dataset, a drug-perturbed, single-cell atlas larger than all public data combined, in collaboration with our friends and partners Parse Biosciences and Ultima Genomics.

Today we are announcing that we will fully open-source Tahoe-100M in February and that we are starting a collaboration with NVIDIA Healthcare to train disease-relevant foundation models of human cell on this and other Tahoe-generated data

Open sourcing a dataset of this magnitude is a momentous step towards creating a more open and collaborative community in biological research, which can ultimately help us design better therapeutics for patients.

Thank you NVIDIA for supporting this historic step. We hope others will follow suit.