Published June 6, 2023
| Version v1
Dataset
Open
Enhancer design manuscript data and models
Authors/Creators
- 1. Research Institute of Molecular Pathology (IMP)
Description
This repository holds the trained DNA accessibility and enhancer activity models, and the data used to train and evaluate the model.
Files for DNA accessibility models:
- Accessibility_models_training_data.tar.gz
- <fold>_sequences_<set>.fa
- FASTA files with DNA sequences of genomic regions from train/val/test sets.
- <fold>_sequences_activity_<set>.txt
- Files with accessibility scores for sequences from train/val/test sets.
- <fold>_sequences_<set>.fa
- Accessibility_model_files.tar.gz
- model.h5 and model.json files for each Results_<fold>_<tissue>_DeepSTARR2_<rep>
- Accessibility_models_test_set_predictions.rds
- RDS object with predictions for accessibility
Files for enhancer activity models:
- EnhancerActivity_model_files.tar.gz
- <fold>_sequences_<set>.fa
- FASTA files with DNA sequences of genomic regions from train/val/test sets.
- <fold>_sequences_activity_<set>.txt
- Files with enhancer activity labels for sequences from train/val/test sets.
- <fold>_sequences_<set>.fa
- EnhancerActivity_model_files.tar.gz
- model.h5, model.json and Model_evaluation.pdf files for each Results_<fold>_<tissue>_<rep>
- EnhancerActivity_models_clean_evaluation_data.rds
- RDS object with sequences used for final evaluation of the models
- EnhancerActivity_models_results_per_tissue_test_set.rds
- RDS object with predictions and results of the enhancer activity models
Files
Files
(3.0 GB)
| Name | Size | Download all |
|---|---|---|
|
md5:d51a1e88ed86201d2e1a5e8fc08f3514
|
129.5 MB | Download |
|
md5:fcd59f932dc3960c587793c255f4744c
|
257.8 MB | Download |
|
md5:fe406913b1a1bcece46a5a2cc2c96b5a
|
2.3 GB | Download |
|
md5:a02bdda4bafe78f472d714b727fe9207
|
109.7 MB | Download |
|
md5:e483a8b3aadbb7c9e7d4eb224515ecc0
|
523.3 kB | Download |
|
md5:36f2c102c646a56df4868325c05ad091
|
1.7 MB | Download |
|
md5:bf54b5d8597d337a4ac75b35c1ecd711
|
168.4 MB | Download |