KosinskiLab / AlphaPulldown

https://doi.org/10.1093/bioinformatics/btac749
GNU General Public License v3.0
176 stars 39 forks source link

Improve resume predictions #355

Closed DimaMolod closed 3 weeks ago

DimaMolod commented 4 weeks ago

Read from pkl and update prediction results in the loop where the calculation of START is made. Don't update prediction_result with sequence and protein if already present in the dict. Read & update from pkl is moved to a separate function for clarity.

dingquanyu commented 4 weeks ago

Variables not defined in _read_from_prediction_result function: multimeric_obejct: https://github.com/KosinskiLab/AlphaPulldown/blob/03d5a8d4770a3c7d503c5f1ad029b6720f497e60/alphapulldown/folding_backend/alphafold_backend.py#L134 processed_feature_dict: https://github.com/KosinskiLab/AlphaPulldown/blob/03d5a8d4770a3c7d503c5f1ad029b6720f497e60/alphapulldown/folding_backend/alphafold_backend.py#L141 model_runner: https://github.com/KosinskiLab/AlphaPulldown/blob/03d5a8d4770a3c7d503c5f1ad029b6720f497e60/alphapulldown/folding_backend/alphafold_backend.py#L144

dingquanyu commented 3 weeks ago

@DimaMolod looks good! Just to double check, all of your result pickles were retained in this version where remove pickle is turned to False by default right?

DimaMolod commented 3 weeks ago

@DimaMolod looks good! Just to double check, all of your result pickles were retained in this version where remove pickle is turned to False by default right?

That's right. Otherwise, it will start predicting anew from model 2 for all finished jobs(complexes)

dingquanyu commented 3 weeks ago

@DimaMolod looks good! Just to double check, all of your result pickles were retained in this version where remove pickle is turned to False by default right?

That's right. Otherwise, it will start predicting anew from model 2 for all finished jobs(complexes)

Great I will merge them then and close issue #352