Code and data to reproduce results of our work HADES at AAAI 2026: https://arxiv.org/abs/2601.11012
conda env create -f environment.yml
conda activate hades
download pdb_cache.zip from the following link and unzip:
https://1drv.ms/u/c/D255A9FC0E665D01/IQAWI0xZNg1GTqwUa3GD0uB_AUK1IUsBiifKo_SzVylcpB8
python main.py --task GB1 --mut_pos 38 39 40 53
python main.py --task PhoQ --mut_pos 283 284 287 288
- check samples files in
resultsfolder after training is done --use_structureoption runs ESMFold to predict structures during training if you skip step 2, which is time consuming and turned off by default.