modifications (ASP0 and GLU0): not working

[tozturkwashu]

WARNING - general - Can't find modification mappings for the modifications ['ASP0']. The following modification mappings are known: {('C-ter',): <vermouth.map_parser.Mapping object at 0x7f6f2b10cbe0>, ('N-ter',): <vermouth.map_parser.Mapping object at 0x7f6f2b1ff7f0>, ('COOH-ter',): <vermouth.map_parser.Mapping object at 0x7f6f2b10cda0>, ('NH2-ter',): <vermouth.map_parser.Mapping object at 0x7f6f2b10cf28>}

It seems like ASP0 and GLU0 are listed as protein_resnames inside aminoacids.ff file but modification mappings file doesn't contain information on them. How can we implement the protonated GLU and ASP models into martinize2?

Thank you!

[pckroon]

The relevant mappings would have to be added to vermouth/data/mappings/. You can use e.g. vermouth/data/mappings/modifications.mapping as example. Unfortunately we don't have documentation for the file format(s) available at the moment.

[fgrunewald]

protonation states can be set by either naming the AAs correctly or by mutating the relevant residues to the protonated counterparts following amber or charmm naming convention.

[lorena-zuzic]

This seems to be an unsolved issue. In martini3001 ff, there are no parameters for either ASP0 and GLU0. They are listed as protein_resnames, but there are no [ moleculetype ] sections for either of the two. With a current installation of Martinize2 (from last week, using pip), I am getting the same error.

Are the parameters for ASP0 (and GLU0) available somewhere? Currently, the only vermouth-implemented ff that has a defined ASP0 are elnedyn22, martini22, and martini22p, which is not helpful for Martini3.

[pckroon]

According to charmm naming convention these are called GLUP and ASPP, rather than GLU0 and ASP0.

[lorena-zuzic]

The issue remains: aminoacids.ff for Martini3 does not contain parameters for either of the two, so ASPP naming (tested below) does not work if the selected force field is martini3001.

WARNING - unmapped-atom - These atoms are not covered by a mapping. Either your mappings don't describe all atoms (bad idea), or, there's no mapping available for all residues. ['4417A-ASPP304:N', '4419A-ASPP304:CA', '4421A-ASPP304:CB', '4424A-ASPP304:CG', '4425A-ASPP304:OD1', '4426A-ASPP304:OD2', '4428A-ASPP304:C', '4429A-ASPP304:O']

[pckroon]

https://github.com/marrink-lab/vermouth-martinize/blob/master/vermouth/data/force_fields/martini3001/aminoacids.ff#L230 https://github.com/marrink-lab/vermouth-martinize/blob/master/vermouth/data/mappings/martini3001/aspp.charmm36.map

What's the output of martinize2 --version?

[lorena-zuzic]

martinize with vermouth 0.7.3

[pckroon]

Ah right, those parameters have not been released to pypi yet, but it's fixed in master. @fgrunewald I forgot if there was anything blocking a release, were we waiting for a PR/feature?

[fgrunewald]

not really but I think we should call it version v1.0.0 and for that I believe we wanted to also fix the modifications? I've no issue with minting a release with the current master.

[pckroon]

Which modifications were that again? Is there already a PR or issue? I admit I lost track...

[lorena-zuzic]

Tried the dev version and works great - thanks a lot!