This directory contains guides on how to use GraphINVENT. If viewing in a browser (recommended), simply click the link to the desired tutorial to view it.
- 0_setting_up_environment : Instructions on how to set up the GraphINVENT virtual environment.
- 1_introduction : A quick introduction to GraphINVENT. Uses the example dataset gdb13_1K to guide new users through Training and Generation jobs in the code.
- 2_using_a_new_dataset : A tutorial on how to use new datasets to train models in GraphINVENT.
- 3_visualizing_molecules : A quick guide on how to visualize grids of molecules using RDKit.
- 4_transfer_learning : A guide on how to use GraphINVENT for transfer learning tasks.
- 5_benchmarking_with_moses : A guide on how to benchmark GraphINVENT models using the MOSES distribution-based benchmarks.
- 6_preprocessing_large_datasets : A guide on how to preprocess large datasets in GraphINVENT.
- 7_reinforcement_learning : A guide on how to fine-tune GraphINVENT models for molecular optimization and de novo design tasks. [TODO]
If a tutorial doesn't exist for something you'd like to do, contact me and I'll be happy to create one (if I think others would benefit from it and I have time). Similarly, if you find and error in a tutorial, please let me know so that I can correct it.
Rocío Mercado