AetherCell

AetherCell

AetherCell is a generative framework for virtual cell perturbation, drug response prediction, and drug repurposing from transcriptomic data.

This Hugging Face repository provides the packaged model weights, the accompanying Python API for local inference, and a compressed archive of Agent Skills required for AI-driven analysis. For full documentation, workflows, benchmarks, and project updates, please refer to the main GitHub repository.

Attribution is mandatory.

Any use of these model weights, their outputs, or any derivative model — including fine-tuned, adapted, retrained, distilled, extended, or otherwise improved versions — in any manuscript, preprint, report, benchmark, presentation, model card, repository, or public release must cite the original AetherCell bioRxiv preprint in accordance with the license terms.

Supported tasks

• Virtual perturbation for drug and gene perturbations
• Drug response prediction for cancer cell lines
• Drug repurposing for disease-oriented candidate ranking

Repository purpose

Use this repository to:

• download AetherCell model assets
• access the Python API for local workflows
• load the weights required by the main AetherCell codebase

Links

• Code & documentation: GitHub - AetherCell
• Processed datasets: Zenodo Record
• Preprint: bioRxiv

Getting started

Please follow the installation and usage instructions in the GitHub repository.
The GitHub repository contains the latest environment setup, inference examples, workflow entry points, and reproducibility resources.

Intended use

AetherCell is intended for:

• non-commercial academic research
• non-commercial scientific evaluation
• local inference and downstream exploratory analysis
• non-commercial fine-tuning, adaptation, or improvement for research purposes

Out-of-scope use

AetherCell and its associated model assets are not intended for:

• clinical diagnosis
• patient stratification
• treatment selection or treatment decision-making
• commercial deployment or commercial services
• redistribution of model weights as standalone assets without permission

Limitations

• Performance may vary across perturbation classes, cell lines, and biological contexts.
• Gene perturbation tasks depend on the availability and quality of perturbation-specific representations.
• Drug repurposing outputs are hypothesis-generating and require downstream experimental validation.
• Use of the model outside the tested research settings may lead to unreliable conclusions.

Responsible use

FOR RESEARCH USE ONLY

This repository and its associated model assets are intended for research use only. They are not validated for clinical use, diagnosis, patient stratification, or treatment decision-making. Any biological or therapeutic hypothesis generated by the system should be independently evaluated and experimentally validated.

Citation

If you use AetherCell in your research, please cite the following preprint:

@article{li2026aethercell,
  title   = {AetherCell: A Generative Engine for Virtual Cell Perturbation and In Vivo Drug Discovery},
  author  = {Li, Wenyuan and Chen, Yang and Peng, Zhaoyi and Xiang, Lei and Wang, Dong and Xie, Zhi},
  journal = {bioRxiv},
  year    = {2026},
  doi     = {10.64898/2026.03.13.710968},
  url     = {https://www.biorxiv.org/content/10.64898/2026.03.13.710968v1}
}

License

This project is distributed under the AetherCell Research License v1.0.

Permitted use

Non-commercial academic research

Non-commercial scientific evaluation

Internal reproduction for research purposes

Fine-tuning, adaptation, or improvement for non-commercial research only

Conditions

Citation of the AetherCell preprint is mandatory for any use of the repository, model, model weights, outputs, or any derived / fine-tuned / adapted / improved model in a publication, preprint, report, benchmark, presentation, or other public disclosure

Any redistributed derivative model must retain this attribution and citation notice

Any modified version must clearly indicate that changes were made

Prohibited use

Commercial use

Clinical or medical decision-making

Redistribution of model weights as standalone assets without permission

Removing or obscuring attribution, provenance, or citation requirements

See LICENSE for full terms.