PlantHelixSeek-open_chromatin

PlantHelixSeek-open_chromatin

Fine-tuned checkpoint of PlantHelixSeek (470M parameters) for three-class classification of open-chromatin state.

Task

• Type: three-class classification of open-chromatin state
• Label: 0, 1, 2 (three chromatin classes)
• Input length: up to 3,000 bp
• Dataset: plant-multi-species-open-chromatin

Dataset and splits

The checkpoint uses the released train.csv, dev.csv and test.csv splits of plant-multi-species-open-chromatin (118,800 / 14,850 / 14,850 sequences). The source of the dataset is described on its dataset card.

Training

Held-out validation metrics

Metrics on the held-out dev.csv split (the split used for checkpoint selection).

Usage

from transformers import AutoModelForSequenceClassification, AutoTokenizer

model = AutoModelForSequenceClassification.from_pretrained(
    "zhangtaolab/PlantHelixSeek-open_chromatin", trust_remote_code=True
)
tokenizer = AutoTokenizer.from_pretrained(
    "zhangtaolab/PlantHelixSeek-open_chromatin", trust_remote_code=True
)

Always pass trust_remote_code=True; the HelixSeek architecture is registered through the Hugging Face auto_map mechanism.

Citation

Wu Y, He Y, Chen L, Liu Y, Wang T, Zhao H, Zhang Y, Zhang T. PlantHelixSeek learns transferable genome representations that capture conserved genome grammar for functional interpretation across plant species. Molecular Plant, 2026. Manuscript MOLECULAR-PLANT-D-26-00999 (DOI to be assigned).

Dataset source: Liu et al., Molecular Plant 2025; Ding et al., GPB 2023, doi:10.1016/j.gpb.2022.10.005.

License

CC-BY-NC 4.0 (research use; contact the authors for commercial licensing).