---
license: apache-2.0
tags:
- biology
size_categories:
- 10K<n<1M
task_categories:
- token-classification
---
# Cancer-associated eccDNA prediction (short)

This dataset contains **Cancer-associated eccDNA prediction (short)** sampled from healthy and cancer tissues.
Each entry contains:
- `sequence`: DNA string (A/T/C/G)
- `label`: 0 = healthy, 1 = cancer

---

## 📁 Folder Structure
<pre>
cancer_eccdna_prediction_short/
├── data/
│   └── cancer_eccdna_prediction_short.csv
├── hf_data/
│   ├── dataset_info.json
│   └── data.arrow
└── README.md
</pre>

---

## 🚀 Quick Usage

<pre><code class="language-python">
from datasets import load_dataset, load_from_disk

# Load from local
dataset = load_from_disk("cancer_eccdna_prediction_short/hf_data")

# Load from Hugging Face Hub (after upload)
dataset = load_dataset("your-username/cancer_eccdna_prediction_short")
</code></pre>

---

## Citation
If you use this dataset, please cite:
<pre><code class="language-python">

@inproceedings{liu2025eccdnamamba,
  title={eccDNAMamba: A Pre-Trained Model for Ultra-Long eccDNA Sequence Analysis},
  author={Zhenke Liu and Jien Li and Ziqi Zhang},
  booktitle={ICML 2025 GenBio Workshop},
  year={2025},
  url={https://openreview.net/forum?id=56xKN7KJjy}
}