If you are adding a new tool, please, add here a short description. Do not forget to properly document your tool. Tests and demos help users to start smoothly.
With create_seq_genomic_dataset function, you can transform a dataset given as a list of intervals (CSV files + yaml metadata file) into a dataset with the actual genomic sequences (TXT files). See the example in demo_notebook.ipynb.
Sometimes we want to include a benchmark that was given to us as full sequences, not genomic intervals. This tool search for perfect matches in a reference and output a dataframe with genomic locations. For example, see human non-tata promoters dataset.
A collention of functions to get info, check or validate a benchmark. Currently, only a limited scope of tools is implemented. For demo, see demo_data_check.ipynb.
Objects wrapping availible datasets, making them directly usable by machine learning libraries. Currently supports pytorch datasets.
Definition of TF and PyTorch models.