Rtp parser - Abandon all hope, ye who review here

[pckroon]

This pile fixes most of the RTP parser. It's not possible to generate the implicit interactions (specifically the dihedrals) using Links, since there sometimes is a selection criterion to decide which of all the options to keep, which also depends on already existing interactions. In addition, because the links that would add these interactions are super symmetric, it quickly becomes a headache.

[fgrunewald]

@pckroon quick question before I review: does this mean the interactions that depend on others don't work then?

[pckroon]

does this mean the interactions that depend on others don't work then?

Sorry, don't understand the question

[fgrunewald]

If I put in an abitrary rtp file, am I guaranteed to get the correct answer with martinize2?

[pckroon]

Aside from testing and termini, yes.

[pckroon]

Here's termini modifications that I think should work, except for a N-term GLY

[ modification ]
C-ter
[ atoms ]
CA {"element": "C"}
C {"element": "C", "replace": {"atype": "CC", "charge": 0.34}}
O {"element": "O", "replace": {"atomname": "OT1", "atype": "OC", "charge": -0.67}}
OT2 {"element": "O", "PTM_atom": true, "atype": "OC", "charge": -0.67}
[ edges ]
CA C
C O
C OT2
[ bonds ]
C OT2 1
C O 1
[ impropers ]
C CA OT2 O 2

[ modification ]
N-ter
[ atoms ]
CA {"element": "C", "replace": {"charge": 0.21, "atype": "CT1"}}
HA {"element": "H", "replace": {"charge": 0.1, "atype": "HB1"}}
N {"element": "N", "replace": {"charge": -0.3, "atype": "NH3"}}
HN {"element": "H", "replace": {"atomname": "H1", "atype": "HC", "charge": 0.33}}
HN2 {"element": "H", "PTM_atom": true, "replace": {"atomname": "H2", "atype": "HC", "charge": 0.33}}
HN3 {"element": "H", "PTM_atom": true, "replace": {"atomname": "H3", "atype": "HC", "charge": 0.33}}
[ edges ] 
N CA
CA HA
N HN
N HN2
N HN3
[ bonds ]
N CA 1
N HN 1
N HN2 1
N HN3 1

[pckroon]

As a sidenote, I tested this on 1UBQ using charmm36-jul2022.ff