---
dataset_info:
  features:
  - name: sequence
    dtype: string
  - name: label
    dtype: int64
  splits:
  - name: train
    num_bytes: 2893542
    num_examples: 3424
  - name: valid
    num_bytes: 740946
    num_examples: 857
  - name: test
    num_bytes: 2594652
    num_examples: 2529
  download_size: 2914341
  dataset_size: 6229140
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---


From this [paper](https://www.sciencedirect.com/science/article/pii/S2405805X25001905). 1 is high expression, 0 is low.