Not sure is this is the right place to report an issue on this Phenix toolbox, but that's what Google returned as the most relevant result.
Anyway, seems like this piece of code does not work as it should on residues with missing atoms:
For example, it produces wrong occupancies (and B-factors) in the N-terminal MET in 2MSI structure. phenix.AmberPrep on the deposited pdb-file returns zero occupancy for the originally present nitrogen atom and non-zero occupancies and B-factors for the previously missing atoms.
According to the code snippet this is expected, since after sorting nitrogen is trimmed by the zip() instead of the missing CG and other atoms. However, one would expect that only missing atoms have zero occupancies by the end of the file preparation. So, some more intelligent mapping needs to be done.
Hi,
Not sure is this is the right place to report an issue on this Phenix toolbox, but that's what Google returned as the most relevant result. Anyway, seems like this piece of code does not work as it should on residues with missing atoms:
https://github.com/phenix-project/amber_adaptbx/blob/78c1a647cbf7eaa324d409746b62f5ba8d6e953e/command_line/AmberPrepClass.py#L351-L360
For example, it produces wrong occupancies (and B-factors) in the N-terminal MET in 2MSI structure.
phenix.AmberPrepon the deposited pdb-file returns zero occupancy for the originally present nitrogen atom and non-zero occupancies and B-factors for the previously missing atoms. According to the code snippet this is expected, since after sorting nitrogen is trimmed by thezip()instead of the missing CG and other atoms. However, one would expect that only missing atoms have zero occupancies by the end of the file preparation. So, some more intelligent mapping needs to be done.Hope, this will help.