Can Martinize2 working on other molecules besides protein?

[hliu56]

Hi

Can Martinize2 produce coarse-grained structures and topologies from the atomistic structure of fatty acids (e.g. Decanoic_acid, Decanol...etc)? Or it's only for protein?

[pckroon]

In principle yes, but you'd need to supply the required data files. The process is effectively the same as adding a new (protein) residue: https://vermouth-martinize.readthedocs.io/en/latest/tutorials/6_adding_residues_links/index.html
2 notes: 1) double check the command line arguments; the defaults there are rather protein focussed. 2) You cannot map atoms from multiple molecules to the same bead. For example, martinize2 cannot map water, since it takes 4 water molecules to build a single bead.

[hliu56]

Thanks!

So it needs to provide force field file and mapping file along with the all atom pdb file?

I wonder if it's able to only use the all atom pdb file to generate the output by using command like this:
martinize2 -f new.pdb -o topol.top -x cg_new.pdb -ff martini3001

Here is an running case in my terminal:
/Martinize % martinize2 -f DECA.pdb -x cg_protein_DECA.pdb -o topol_DECA.top

It will encounter the below issue after running 1 hour and a half:
INFO - general - Read 1 molecules from PDB file DECA.pdb
INFO - step - Guessing the bonds.
INFO - general - 1 molecules after guessing bonds
WARNING - general - Residue specified by "cter" for mutation "C-ter" not found
INFO - step - Repairing the graph.
zsh: killed martinize2 -f DECA.pdb -x cg_protein_DECA.pdb -o topol_DECA.top

Here is the all atom pdb file that I used:
DECA.pdb.zip

[pckroon]

So it needs to provide force field file and mapping file along with the all atom pdb file?

Yes

I wonder if it's able to only use the all atom pdb file to generate the output by using command like this

No. You need to tell martinize2 what the resulting CG parameters should be for each residue, and you need to tell it how AA atoms correspond to CG beads

You'll also need to do something with termini, since martinize2 by default assumes you're working with proteins (TODO). You can tell it to not put termini by adding -cter none -nter none.