You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I notice in the processed_files subdirs of the DockGen dataset that there are several ligand files. Which of these was used for comparing the different methods? I'd assume just the one with the same filename stem as the prepared protein file, but in that case I think there's a mistake in (at least) 6wyz_1_DGL_1. The processed protein file contains two protein chains with the same fold. There are 4 ligand files of which one binds to chain C, one to D, and 2 are floating out in space. The 'right' answer (ie the one which matches the protein_processed filename, DGL_1, is the one that binds chain D. But it's possible for a docking algorithm to get the right pose but in the wrong chain.
The text was updated successfully, but these errors were encountered:
I notice in the processed_files subdirs of the DockGen dataset that there are several ligand files. Which of these was used for comparing the different methods? I'd assume just the one with the same filename stem as the prepared protein file, but in that case I think there's a mistake in (at least) 6wyz_1_DGL_1. The processed protein file contains two protein chains with the same fold. There are 4 ligand files of which one binds to chain C, one to D, and 2 are floating out in space. The 'right' answer (ie the one which matches the protein_processed filename, DGL_1, is the one that binds chain D. But it's possible for a docking algorithm to get the right pose but in the wrong chain.
The text was updated successfully, but these errors were encountered: