# An open foundation model of the cancer cell trained on perturbation data

Source: https://onco.cc/ideas/idea-data-open-cell-foundation-model/  
OnCo record `idea-data-open-cell-foundation-model` (Idea). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Build a shared, openly available AI model that has learned how cancer cells respond to genetic and drug perturbations, so any lab can predict what a new drug or combination might do.

## Summary

Single-cell perturbation atlases, CRISPR screens (DepMap), drug-response datasets and proteomics now exist at scale, but models trained on them are mostly proprietary or single-lab. The proposal is a pre-competitive, openly licensed foundation model of the cancer cell (transcriptomic and proteomic state under perturbation) trained on pooled public and consortium data with open weights, evaluated on held-out perturbations and prospective wet-lab validation, in the way AlphaFold became shared infrastructure for structure. The Chan Zuckerberg Initiative's virtual cell work and the Arc Institute's efforts are precedents.

## Fields

- Kind: Idea
- Last checked: 2026-09-08
- Hypothesis: An open cell model will predict the transcriptional response to unseen drug and gene perturbations in unseen cell lines with accuracy sufficient to prioritise combinations, and prospectively validated predictions will yield synergistic combinations at a rate several times higher than random screening.
- Rationale: AlphaFold showed that a shared open model on curated public data can lift an entire field; perturbation biology now has the data volume and benchmark structure to attempt the same.
- Proposed test: Train on public perturbation data with a held-out set of drugs and cell lines; test the top 100 predicted synergistic combinations in wet-lab screens against 100 random combinations; report hit rates.
- Maturity: preclinical-evidence
- Actor: philanthropy

## Sources

- DepMap: https://depmap.org/portal/
- CZI virtual cell: https://virtualcellmodels.cziscience.com/

## Connected records

- collections: [CZ CELLxGENE / Human Cell Atlas](https://onco.cc/collections/cellxgene-hca/), [DepMap (Cancer Dependency Map)](https://onco.cc/collections/depmap/)
- fronts: [AI & Computation](https://onco.cc/fronts/ai-computation/), [Drug Discovery Platforms](https://onco.cc/fronts/drug-discovery/)
- technologies: [AI-driven drug & target discovery](https://onco.cc/technologies/ai-drug-design/), [CRISPR functional genomics](https://onco.cc/technologies/crispr-screens/), [Single-cell & spatial profiling](https://onco.cc/technologies/single-cell-spatial/)
- institutions: [Broad Institute of MIT and Harvard](https://onco.cc/institutions/broad-institute/)
- bottlenecks: [AI that is built but not validated or deployed](https://onco.cc/bottlenecks/b-ai-validation/), [Lab models that fail to predict what happens in patients](https://onco.cc/bottlenecks/b-preclinical-models/), [Too many combinations to test](https://onco.cc/bottlenecks/b-combination-space/)
- ideas: [In silico trials to prioritise combinations, scored against later real trials](https://onco.cc/ideas/idea-data-in-silico-trials-calibrated/)
- terms: [Cancer AI vocabulary (CanSim terms map)](https://onco.cc/terms/cancer-ai-vocabulary/), [Virtual cell models and in-silico perturbation screens](https://onco.cc/terms/virtual-cell-models/)

---
JSON: https://onco.cc/api/v1/entities/idea-data-open-cell-foundation-model.json