About the training data

[Qmi3]

Great Work! It seems that ModelAngelo organizes a training set with good quality. Do you have plans to release them? I think it's a great resources for the computational CryoEM community.

[jamaliki]

Hi @Qmi3 ,

Sorry for the late reply! The original dataset with all the processing already done etc is too large for us to be able to release, due to the size of the maps associated with it.

What I can offer instead is the following list of PDB ids used for training. I can also (if there is interest) provide some cleaned scripts for the PDB parsing, some constant shifts required to register cryo-EM maps with these models, and some pointers for processing the maps.

Please find the attached list of PDB ids: model_angelo_train_pdbs.txt

Best, Kiarash.