RNA3DB-2D dataset

This release provides machine learning–ready RNA secondary structure datasets in pickle format, generated using the RNA3DB-2D-Structures pipeline.

Contents
RNA3DB-2D-pickles-less600.tar.gz
train.pickle: trainsetvalid.pickle: validsettest.pickle: testsetThese files were generated by the script 07_bpseq_to_pickle.py from the RNA3DB-2D-Structures project.

Important:
Only RNA sequences with length ≤ 600 nt are included (longer sequences were filtered out by default).
To generate datasets with a different sequence length limit, please modify the --max_length parameter in 07_bpseq_to_pickle.py(or the PICKLE_MAX_LENGTH variable in run_pipeline.sh) and rerun the script.
For details on the format and how to use these files, please refer to the RNA3DB-2D-Structures documentation.