virtual-cell
ai4sci-virtual-cell-zeroshot-data
ai4sci virtual-cell-zeroshot — release v1
Prepared data for the virtual-cell-zeroshot
task: predict single-cell CRISPRi knockdown responses in cellular contexts a model has never seen
perturbed (the Arc Virtual Cell Challenge 2026 zero-shot setting), rebuilt from public data.
dev/ what the agent sees (mount read-only at /workspace/data)
train/{k562,jurkat,hct116,hek293t}/ cells.h5ad, pseudobulk.h5ad, se_embeddings.npy, dev_split/
test/{ctx_near,ctx_mid… See the full description on the dataset page: https://huggingface.co/datasets/sunweiwei/ai4sci-virtual-cell-zeroshot-data.arc-institute-virtual-cell-dataset
ARC Institute Virtual Cell Challenge
Please check out the official website for the challenge rules and deadlines.
About
For this challenge, single-cell functional genomics was used to generate approximately 300,000 single-cell RNA-seq profiles by silencing 300 carefully selected genes using CRISPR interference (CRISPRi). 10x Genomics GEM-X Flex and Illumina sequencing were used to obtain single-cell gene expression profiles. The data are split into three groups for the… See the full description on the dataset page: https://huggingface.co/datasets/cyrilzakka/arc-institute-virtual-cell-dataset.SE600M-embedding
Instruction from Wanwan Feng
Step 1 — Generate Embeddings for All Input Files
This step shows the creation of this SE600M-embedding dataset from the existing training dataset provided by the Arc Institute for the Virtual Cell Challenge.
import os
os.environ['MPLBACKEND'] = 'Agg'
# 1. Files to process
FILENAMES_TO_PROCESS = [
"competition_train.h5",
"jurkat.h5",
"k562.h5",
"k562_gwps.h5",
"rpe1.h5",
"competition_val_template.h5ad"… See the full description on the dataset page: https://huggingface.co/datasets/VirtualCell2025/SE600M-embedding.2025_Virtual_Cell_Challenge_Test_DataInfoton_Virtual_Cell_Long_Covid_SARS-COV-2datasets