---
configs:
- config_name: EMP_H3
  data_files:
  - split: train
    path: EMP_H3/EMP_H3_train.csv.gz
  - split: validation
    path: EMP_H3/EMP_H3_val.csv.gz
  - split: test
    path: EMP_H3/EMP_H3_test.csv.gz

- config_name: EMP_H3K14ac
  data_files:
  - split: train
    path: EMP_H3K14ac/EMP_H3K14ac_train.csv.gz
  - split: validation
    path: EMP_H3K14ac/EMP_H3K14ac_val.csv.gz
  - split: test
    path: EMP_H3K14ac/EMP_H3K14ac_test.csv.gz

- config_name: EMP_H3K36me3
  data_files:
  - split: train
    path: EMP_H3K36me3/EMP_H3K36me3_train.csv.gz
  - split: validation
    path: EMP_H3K36me3/EMP_H3K36me3_val.csv.gz
  - split: test
    path: EMP_H3K36me3/EMP_H3K36me3_test.csv.gz

- config_name: EMP_H3K4me1
  data_files:
  - split: train
    path: EMP_H3K4me1/EMP_H3K4me1_train.csv.gz
  - split: validation
    path: EMP_H3K4me1/EMP_H3K4me1_val.csv.gz
  - split: test
    path: EMP_H3K4me1/EMP_H3K4me1_test.csv.gz

- config_name: EMP_H3K4me2
  data_files:
  - split: train
    path: EMP_H3K4me2/EMP_H3K4me2_train.csv.gz
  - split: validation
    path: EMP_H3K4me2/EMP_H3K4me2_val.csv.gz
  - split: test
    path: EMP_H3K4me2/EMP_H3K4me2_test.csv.gz

- config_name: EMP_H3K4me3
  data_files:
  - split: train
    path: EMP_H3K4me3/EMP_H3K4me3_train.csv.gz
  - split: validation
    path: EMP_H3K4me3/EMP_H3K4me3_val.csv.gz
  - split: test
    path: EMP_H3K4me3/EMP_H3K4me3_test.csv.gz

- config_name: EMP_H3K79me3
  data_files:
  - split: train
    path: EMP_H3K79me3/EMP_H3K79me3_train.csv.gz
  - split: validation
    path: EMP_H3K79me3/EMP_H3K79me3_val.csv.gz
  - split: test
    path: EMP_H3K79me3/EMP_H3K79me3_test.csv.gz

- config_name: EMP_H3K9ac
  data_files:
  - split: train
    path: EMP_H3K9ac/EMP_H3K9ac_train.csv.gz
  - split: validation
    path: EMP_H3K9ac/EMP_H3K9ac_val.csv.gz
  - split: test
    path: EMP_H3K9ac/EMP_H3K9ac_test.csv.gz

- config_name: EMP_H4
  data_files:
  - split: train
    path: EMP_H4/EMP_H4_train.csv.gz
  - split: validation
    path: EMP_H4/EMP_H4_val.csv.gz
  - split: test
    path: EMP_H4/EMP_H4_test.csv.gz

- config_name: EMP_H4ac
  data_files:
  - split: train
    path: EMP_H4ac/EMP_H4ac_train.csv.gz
  - split: validation
    path: EMP_H4ac/EMP_H4ac_val.csv.gz
  - split: test
    path: EMP_H4ac/EMP_H4ac_test.csv.gz

- config_name: EPI_GM12878
  data_files:
  - split: train
    path: EPI_GM12878/EPI_GM12878_train.csv.gz
  - split: validation
    path: EPI_GM12878/EPI_GM12878_val.csv.gz
  - split: test
    path: EPI_GM12878/EPI_GM12878_test.csv.gz

- config_name: EPI_HUVEC
  data_files:
  - split: train
    path: EPI_HUVEC/EPI_HUVEC_train.csv.gz
  - split: validation
    path: EPI_HUVEC/EPI_HUVEC_val.csv.gz
  - split: test
    path: EPI_HUVEC/EPI_HUVEC_test.csv.gz

- config_name: EPI_HeLa-S3
  data_files:
  - split: train
    path: EPI_HeLa-S3/EPI_HeLa-S3_train.csv.gz
  - split: validation
    path: EPI_HeLa-S3/EPI_HeLa-S3_val.csv.gz
  - split: test
    path: EPI_HeLa-S3/EPI_HeLa-S3_test.csv.gz

- config_name: EPI_IMR90
  data_files:
  - split: train
    path: EPI_IMR90/EPI_IMR90_train.csv.gz
  - split: validation
    path: EPI_IMR90/EPI_IMR90_val.csv.gz
  - split: test
    path: EPI_IMR90/EPI_IMR90_test.csv.gz

- config_name: EPI_K562
  data_files:
  - split: train
    path: EPI_K562/EPI_K562_train.csv.gz
  - split: validation
    path: EPI_K562/EPI_K562_val.csv.gz
  - split: test
    path: EPI_K562/EPI_K562_test.csv.gz

- config_name: EPI_NHEK
  data_files:
  - split: train
    path: EPI_NHEK/EPI_NHEK_train.csv.gz
  - split: validation
    path: EPI_NHEK/EPI_NHEK_val.csv.gz
  - split: test
    path: EPI_NHEK/EPI_NHEK_test.csv.gz

- config_name: mouse_0
  data_files:
  - split: train
    path: mouse_0/mouse_0_train.csv.gz
  - split: validation
    path: mouse_0/mouse_0_val.csv.gz
  - split: test
    path: mouse_0/mouse_0_test.csv.gz

- config_name: mouse_1
  data_files:
  - split: train
    path: mouse_1/mouse_1_train.csv.gz
  - split: validation
    path: mouse_1/mouse_1_val.csv.gz
  - split: test
    path: mouse_1/mouse_1_test.csv.gz

- config_name: mouse_2
  data_files:
  - split: train
    path: mouse_2/mouse_2_train.csv.gz
  - split: validation
    path: mouse_2/mouse_2_val.csv.gz
  - split: test
    path: mouse_2/mouse_2_test.csv.gz

- config_name: mouse_3
  data_files:
  - split: train
    path: mouse_3/mouse_3_train.csv.gz
  - split: validation
    path: mouse_3/mouse_3_val.csv.gz
  - split: test
    path: mouse_3/mouse_3_test.csv.gz

- config_name: mouse_4
  data_files:
  - split: train
    path: mouse_4/mouse_4_train.csv.gz
  - split: validation
    path: mouse_4/mouse_4_val.csv.gz
  - split: test
    path: mouse_4/mouse_4_test.csv.gz

- config_name: prom_300_all
  data_files:
  - split: train
    path: prom_300_all/prom_300_all_train.csv.gz
  - split: validation
    path: prom_300_all/prom_300_all_val.csv.gz
  - split: test
    path: prom_300_all/prom_300_all_test.csv.gz

- config_name: prom_300_notata
  data_files:
  - split: train
    path: prom_300_notata/prom_300_notata_train.csv.gz
  - split: validation
    path: prom_300_notata/prom_300_notata_val.csv.gz
  - split: test
    path: prom_300_notata/prom_300_notata_test.csv.gz

- config_name: prom_300_tata
  data_files:
  - split: train
    path: prom_300_tata/prom_300_tata_train.csv.gz
  - split: validation
    path: prom_300_tata/prom_300_tata_val.csv.gz
  - split: test
    path: prom_300_tata/prom_300_tata_test.csv.gz

- config_name: prom_core_all
  data_files:
  - split: train
    path: prom_core_all/prom_core_all_train.csv.gz
  - split: validation
    path: prom_core_all/prom_core_all_val.csv.gz
  - split: test
    path: prom_core_all/prom_core_all_test.csv.gz

- config_name: prom_core_notata
  data_files:
  - split: train
    path: prom_core_notata/prom_core_notata_train.csv.gz
  - split: validation
    path: prom_core_notata/prom_core_notata_val.csv.gz
  - split: test
    path: prom_core_notata/prom_core_notata_test.csv.gz

- config_name: prom_core_tata
  data_files:
  - split: train
    path: prom_core_tata/prom_core_tata_train.csv.gz
  - split: validation
    path: prom_core_tata/prom_core_tata_val.csv.gz
  - split: test
    path: prom_core_tata/prom_core_tata_test.csv.gz

- config_name: tf_0
  data_files:
  - split: train
    path: tf_0/tf_0_train.csv.gz
  - split: validation
    path: tf_0/tf_0_val.csv.gz
  - split: test
    path: tf_0/tf_0_test.csv.gz

- config_name: tf_1
  data_files:
  - split: train
    path: tf_1/tf_1_train.csv.gz
  - split: validation
    path: tf_1/tf_1_val.csv.gz
  - split: test
    path: tf_1/tf_1_test.csv.gz

- config_name: tf_2
  data_files:
  - split: train
    path: tf_2/tf_2_train.csv.gz
  - split: validation
    path: tf_2/tf_2_val.csv.gz
  - split: test
    path: tf_2/tf_2_test.csv.gz

- config_name: tf_3
  data_files:
  - split: train
    path: tf_3/tf_3_train.csv.gz
  - split: validation
    path: tf_3/tf_3_val.csv.gz
  - split: test
    path: tf_3/tf_3_test.csv.gz

- config_name: tf_4
  data_files:
  - split: train
    path: tf_4/tf_4_train.csv.gz
  - split: validation
    path: tf_4/tf_4_val.csv.gz
  - split: test
    path: tf_4/tf_4_test.csv.gz
---

# GUE_v2

## Overview

[GUE_v2](https://arxiv.org/pdf/2306.15006) is a collection of multi-species genome classification datasets. This respository includes datasets from GUE (v2) that have been filtered by our criteria to be included in our quality-curated genomic benchmark datasets.

## Dataset structure

The repository is organized as follows:
```
GUE_v2/
├── README.md
├── EMP_H3/
│   ├── EMP_H3_train.csv.gz
│   ├── EMP_H3_val.csv.gz
│   └── EMP_H3_test.csv.gz
├── EMP_H3K14ac/
│   ├── EMP_H3K14ac_train.csv.gz
│   ├── EMP_H3K14ac_val.csv.gz
│   └── EMP_H3K14ac_test.csv.gz
├── EMP_H3K36me3/
│   ├── EMP_H3K36me3_train.csv.gz
│   ├── EMP_H3K36me3_val.csv.gz
│   └── EMP_H3K36me3_test.csv.gz
├── EMP_H3K4me1/
│   ├── EMP_H3K4me1_train.csv.gz
│   ├── EMP_H3K4me1_val.csv.gz
│   └── EMP_H3K4me1_test.csv.gz
├── EMP_H3K4me2/
│   ├── EMP_H3K4me2_train.csv.gz
│   ├── EMP_H3K4me2_val.csv.gz
│   └── EMP_H3K4me2_test.csv.gz
├── EMP_H3K4me3/
│   ├── EMP_H3K4me3_train.csv.gz
│   ├── EMP_H3K4me3_val.csv.gz
│   └── EMP_H3K4me3_test.csv.gz
├── EMP_H3K79me3/
│   ├── EMP_H3K79me3_train.csv.gz
│   ├── EMP_H3K79me3_val.csv.gz
│   └── EMP_H3K79me3_test.csv.gz
├── EMP_H3K9ac/
│   ├── EMP_H3K9ac_train.csv.gz
│   ├── EMP_H3K9ac_val.csv.gz
│   └── EMP_H3K9ac_test.csv.gz
├── EMP_H4/
│   ├── EMP_H4_train.csv.gz
│   ├── EMP_H4_val.csv.gz
│   └── EMP_H4_test.csv.gz
├── EMP_H4ac/
│   ├── EMP_H4ac_train.csv.gz
│   ├── EMP_H4ac_val.csv.gz
│   └── EMP_H4ac_test.csv.gz
├── EPI_GM12878/
│   ├── EPI_GM12878_train.csv.gz
│   ├── EPI_GM12878_val.csv.gz
│   └── EPI_GM12878_test.csv.gz
├── EPI_HUVEC/
│   ├── EPI_HUVEC_train.csv.gz
│   ├── EPI_HUVEC_val.csv.gz
│   └── EPI_HUVEC_test.csv.gz
├── EPI_HeLa-S3/
│   ├── EPI_HeLa-S3_train.csv.gz
│   ├── EPI_HeLa-S3_val.csv.gz
│   └── EPI_HeLa-S3_test.csv.gz
├── EPI_IMR90/
│   ├── EPI_IMR90_train.csv.gz
│   ├── EPI_IMR90_val.csv.gz
│   └── EPI_IMR90_test.csv.gz
├── EPI_K562/
│   ├── EPI_K562_train.csv.gz
│   ├── EPI_K562_val.csv.gz
│   └── EPI_K562_test.csv.gz
├── EPI_NHEK/
│   ├── EPI_NHEK_train.csv.gz
│   ├── EPI_NHEK_val.csv.gz
│   └── EPI_NHEK_test.csv.gz
├── mouse_0/
│   ├── mouse_0_train.csv.gz
│   ├── mouse_0_val.csv.gz
│   └── mouse_0_test.csv.gz
├── mouse_1/
│   ├── mouse_1_train.csv.gz
│   ├── mouse_1_val.csv.gz
│   └── mouse_1_test.csv.gz
├── mouse_2/
│   ├── mouse_2_train.csv.gz
│   ├── mouse_2_val.csv.gz
│   └── mouse_2_test.csv.gz
├── mouse_3/
│   ├── mouse_3_train.csv.gz
│   ├── mouse_3_val.csv.gz
│   └── mouse_3_test.csv.gz
├── mouse_4/
│   ├── mouse_4_train.csv.gz
│   ├── mouse_4_val.csv.gz
│   └── mouse_4_test.csv.gz
├── prom_300_all/
│   ├── prom_300_all_train.csv.gz
│   ├── prom_300_all_val.csv.gz
│   └── prom_300_all_test.csv.gz
├── prom_300_notata/
│   ├── prom_300_notata_train.csv.gz
│   ├── prom_300_notata_val.csv.gz
│   └── prom_300_notata_test.csv.gz
├── prom_300_tata/
│   ├── prom_300_tata_train.csv.gz
│   ├── prom_300_tata_val.csv.gz
│   └── prom_300_tata_test.csv.gz
├── prom_core_all/
│   ├── prom_core_all_train.csv.gz
│   ├── prom_core_all_val.csv.gz
│   └── prom_core_all_test.csv.gz
├── prom_core_notata/
│   ├── prom_core_notata_train.csv.gz
│   ├── prom_core_notata_val.csv.gz
│   └── prom_core_notata_test.csv.gz
├── prom_core_tata/
│   ├── prom_core_tata_train.csv.gz
│   ├── prom_core_tata_val.csv.gz
│   └── prom_core_tata_test.csv.gz
├── tf_0/
│   ├── tf_0_train.csv.gz
│   ├── tf_0_val.csv.gz
│   └── tf_0_test.csv.gz
├── tf_1/
│   ├── tf_1_train.csv.gz
│   ├── tf_1_val.csv.gz
│   └── tf_1_test.csv.gz
├── tf_2/
│   ├── tf_2_train.csv.gz
│   ├── tf_2_val.csv.gz
│   └── tf_2_test.csv.gz
├── tf_3/
│   ├── tf_3_train.csv.gz
│   ├── tf_3_val.csv.gz
│   └── tf_3_test.csv.gz
└── tf_4/
    ├── tf_4_train.csv.gz
    ├── tf_4_val.csv.gz
    └── tf_4_test.csv.gz
```