JMK parser fix to account for xmipp pseudoatom writing changes

[jamesmkrieger]

Accounts for I2PC/xmipp3#664

With both changes, the TestHEMNMA_1 works fine

[MohamadHarastani]

I would ask Remi to verify since he wrote this file. I will approve it from my side.

[jamesmkrieger]

Thanks, Mohamad.

Remi, can you also look at the changes on the xmipp side, please?

[jamesmkrieger]

We're still thinking further about this on the xmipp side, so I'm converting this to draft

[mms29]

Hi James,
Sorry for the late reply I was on holidays.
The PDB parser is based on the specification found here : https://www.cgl.ucsf.edu/chimera/docs/UsersGuide/tutorials/pdbintro.html
Your modifications do not follow this specification so I assume it will not work for some PDBs (Although it might work in some cases)
It would help for me to have some background on the issue, what was the error of TestHEMNMA_1 you mentioned ?

@MohamadHarastani
I haven't used this pdb parser (I use continuousflex/protocols/utilities/pdb_handler)
I don't know who wrote this but it is similar to mine for reading and writing

[jamesmkrieger]

There is no error with the tests that I’m fixing. Rather, I’m mentioning the test to say that I’m not breaking anything.

The idea here is indeed to aim towards having a pseudoatom pdb file that actually follows the pdb format. This parser that I edited is the specific one for that, which in theory could probably be retired. David Strelak and I were thinking we could do the same on the Xmipp side.

[mms29]

I believe PDB file with atoms or pseudoatoms should follow the exact same specification, maybe only the residue name will be specific to pseudoatoms, but the rest should be the same. I think this is how it is done so far in continuousflex. I don't understand the problem you are having and why you should modify this PDB reader to pseudoatoms pdb files.
Also, I believe this function is referred from the NMA protocol and is not dedicated to pseudoatoms files, so by doing this modification it might break the NMA protocol for some atomic PDBs?

[jamesmkrieger]

They should be but they aren’t actually. I’ll have a look at this all more carefully