Virtual Cell Challenge
The competition, hosted by Arc Institute at virtualcellchallenge.org and sponsored by NVIDIA, 10x Genomics, and Ultima Genomics, is focused on accelerating progress in AI modeling of biology. Entrants will create models that will predict effects of single gene perturbations in cells. With a grand prize worth $100,000, the challenge is part of Arc's efforts to inspire the creation of high-quality datasets and help establish standardized benchmarks for virtual cell modeling.
Predictions are submitted as .vcc files, produced by the vcc command-line tool. The .vcc format is a wrapper around an AnnData h5ad file that ensures the submission meets the requirements of the Challenge. You build one h5ad covering all three cell contexts — not one file per context — and package it with vcc prep, which validates the format locally so problems surface in seconds rather than after a multi-gigabyte upload: vcc prep prediction.h5ad -g gene_names.csv --perts pert_counts.csv -o prediction.vccvcc submit prediction.vcc -m "my model v3" --wait If you use a coding agent (Claude…
related reading
- Virtual Cell Challengevirtualcellchallenge.org
- Virtual Cell Challengevirtualcellchallenge.org
- Virtual Cell Challenge 2025 Wrap-Up: Winners and Reflections | Arc Institutearcinstitute.org
- Arc Institute’s first virtual cell model: <span style="font-variant: small-caps">S<span style="font-weight: bolder">tate</span></span> | Arc Institutearcinstitute.org
- Arc Virtual Cell Challenge: A Primerfleetwood.dev
- Towards building a World Model to simulate perturbation-induced cellular dynamics by AlphaCellbiorxiv.org
- Cellular Agriculture AI Librarytucca-cellag.github.io
- ML Contestsmlcontests.com
- Kaggle Solutionsfarid.one
- CodaLab Worksheetsworksheets.codalab.org
- AI Model & API Providers Analysis | Artificial Analysisartificialanalysis.ai
- Tahoe 100M: The World's Largest Single-Cell Dataset, Open-Sourced as the Inaugural Contribution to Arc Institute's New Virtual Cell Atlas | Tahoetahoebio.ai