Pages

Friday, 31 May 2019

A note on the Linux PyMOL C01 atom oddity

This weird bug has been haunting me for ages. The PyMOL 1.8 (not 2 in Win or Mac) and Linux PyMOL 2 builder creates residues with a Cα called C01 as opposed to CA. If any operation is done to these (e.g. Rosetta Relax), they will be discarded during the reading of the file. That is, they will not be fixed and worse if Rosetta Remodel is used, it will assume that the residue never existed, because Remodel does not understand PDB numbering annoyingly. Simply substituting all 'C01' to 'CA' fixes the problem.

The installer version for Win and Mac is 2.x, while the Conda install version (the open source version) is 1.8, so this bug only appear in Linux or systems where PyMol was installed with Conda. The residue build and mutagenesis wizard does it both in the command line and in the GUI. It is clear if using the GUI that something is wrong as the backbone will be absent even if the residues are there.
It is a super corner case but hopefully this note helps someone...

No comments:

Post a Comment