Page 1 of 1

Writing an amber force field which fits a branched polymer!

Posted: Fri Oct 14, 2022 8:46 am
by elsaid
I am developing an amber force field for a branched polymer, but so far it did not work.
I am looking for any source helping me to understand <HarmonicBondForce>,
<HarmonicAngleForce> and <PeriodicTorsionForce>.

Best,
Elsaid

Re: Writing an amber force field which fits a branched polymer!

Posted: Fri Oct 14, 2022 9:56 am
by peastman
Is there something specific you want to understand? If you're asking about the syntax of the XML file, that's described in http://docs.openmm.org/latest/userguide ... g_ffs.html. If you're asking about the functional forms of the different forces, see http://docs.openmm.org/latest/userguide ... orces.html.

Re: Writing an amber force field which fits a branched polymer!

Posted: Fri Oct 14, 2022 2:55 pm
by elsaid
I successfully add the force block for a linear one, and it works. But I can not do a modified xml file to get the pdb file of the branched polymer get accepted by the program.
I think I have to understand deeply how bonds types are written there.

For that chain:(just an example )
(CHO2)(CH3)(CH3)
Every methy group is attached to an O from the first residue
I need to have an amber force field to the above (force blocks especially)
Thanks.

Re: Writing an amber force field which fits a branched polymer!

Posted: Wed Oct 19, 2022 5:37 am
by elsaid
What is the difference between the amber force field for linear and branched structures?

Re: Writing an amber force field which fits a branched polymer!

Posted: Sat Oct 22, 2022 9:15 am
by elsaid
The pdb file has 2 types of residues "COR and EGH" in 1 chain "A"
I define the residues like:
<Residue>
<Residue name="COR" >
C................
</Residue>
<Residue>
<Residue name="EGH" >
C................
</Residue>

Then I am getting the error:
anaconda3/envs/opki/lib/python3.10/site-packages/openmm/app/forcefield.py", line 295, in loadFile
resName = resname_prefix+residue.attrib['name']

KeyError: 'name'
there is something wrong?