# PromoterAtlas Dataset

This repository contains the training data used for the PromoterAtlas model, a transformer-based deep learning model for bacterial regulatory sequence analysis.  

### Dataset Description
Content: ~9 million regulatory sequences from over 3,000 gammaproteobacterial species. The sequences were extracted from GenBank files of gammaproteobacterial genomes.  
Format: HDF5 file (sequence_dataset.h5)  
Size: Approximately 2.3 GB  

### Usage
This dataset is primarily intended for training the PromoterAtlas-base model. For access to the different PromoterAtlas models and code, please visit: https://github.com/LucasCoppens/PromoterAtlas