virtual-cell

The Arc Institute is a well-funded nonprofit research institute building 'virtual cell' AI models and the datasets to train them.

Arc's growing library of cell data, the fuel for virtual cell models.

Turns a cell's gene expression into a sentence so a normal language model can reason about it; a 27-billion-parameter version proposed a cancer immunotherapy idea that was confirmed in the lab.

CellFM is an 800-million-parameter single-cell model trained on 100 million human cells.

Funds and builds the Human Cell Atlas infrastructure and the CZI virtual cell models (TranscriptFormer), plus one of the largest nonprofit GPU clusters for biology.

GEARS is a graph model predicting the effect of gene knockouts; the perturbation benchmarks around it showed how hard the problem is.

Geneformer is a transformer trained on about 30 million single cells that encodes each cell as a ranked list of its genes, so deleting a gene in silico shows which genes matter in a disease. It was the first single-cell foundation model in general use, though benchmarks find only modest gains over linear baselines on some tasks.

Uses text embeddings of gene descriptions from a general LLM to represent cells, and performs surprisingly well.

Nicheformer is a model trained on both dissociated and spatial data so it learns how a cell's neighbourhood shapes it.

scFoundation is a 100-million-parameter model trained on 50 million cells, from China's BioMap.

A GPT-style model for single-cell data that predicts cell types, perturbation responses, and gene networks.

Predicts how cells will respond to a drug or gene knockout, trained on over 100 million perturbed cells.

Tahoe-100M is the biggest single-cell dataset ever released, built to teach AI how cancer cells respond to drugs.

CZI's open cross-species cell models and a reasoning model trained on them.

Universal Cell Embedding maps any cell from any species into one shared space without retraining.

Produced Tahoe-100M, the largest single-cell drug-perturbation atlas, and trains models on it.

Home of the Cell2Sentence single-cell language models built with Google, and an NCI-designated comprehensive cancer centre.