search

openmm / openmm

OpenMM is a toolkit for molecular simulation using high performance GPU code.
1.48k stars 519 forks source link

How to handle attached molecules to residues in OpenMM #3353

Open hector-s-m opened 2 years ago

hector-s-m commented 2 years ago

(Heads up, very amateur coder and new to MD)

I am trying to run a simulation in which I covalently attached two fluorophores molecules (Alexa 555 and Atto 647N) to the crystal structure of my protein of interest. I built the fluorophores with the 'build' tool of Pymol, and saved the molecule with the CONNECT info (allowing multiple CONNECT for double bonds). If I open the .pdb file as text, I see the coordinates of each atom belonging to the same residue that I covalently linked the molecule to. Will this allow me to run a simulation using OpenMM? What force field should I use? I used:

forcefield = ForceField('amber99sb.xml', 'tip3p.xml')
system = forcefield.createSystem(pdb.topology, nonbondedMethod=CutoffNonPeriodic)

After playing around for a bit, I obtained the following error: "No template found for residue 1". Note that residue 1 is not the residue to which I attached the fluorophore molecule.

Attached is the PDB file in case someone wants/can take a look in closer detail. Appreciate the help forehand. Best, Hector LipA-fluorophores.pdb.zip

peastman commented 2 years ago

Can OpenFF be used to parametrize fluorophores attached to a protein?

jchodera commented 2 years ago

There is currently no way to automatically generate parameters for covalently attached small molecules with Amber, but CHARMM-GUI does provide support for many of these fluorophores: https://www.charmm-gui.org/

For the moment, I'd suggest using CHARMM-GUI to set up your system and then run it with OpenMM.

Soon, the Open Force Field Initiative will provide a way to consistently generate parameters for small molecules and biopolymers, including covalent small molecules like this---currently slated for 2022.

A (highly) experimental way to do this right now is using espaloma.

aizvorski commented 2 years ago

@hector-s-m Have a look at this Amber tutorial: http://ambermd.org/tutorials/basic/tutorial5/index.php "Simulating the Green Fluorescent Protein and Building a Modified Amino Acid Residue"

You can prepare your system exactly as described in there, just use a modified residue consisting of your fluorophore plus the residue it is linked to. Once you have the Amber parm7 / rst7 files made by tleap, you can load them and run the simulations in OpenMM.

aizvorski commented 2 years ago

@hector-s-m I was just browsing the Amber version 21 manual http://ambermd.org/doc12/Amber21.pdf and noticed this section: "3.10. Fluorescent dyes: AMBER-DYES in AMBER force field files". There is an example of making a system with a covalent dye using tleap in there, so it seems what you want to do is even easier now. Start by installing the latest ambertools with conda install -c conda-forge ambertools=21 then follow the manual.

Good luck, and please do post some instructions here when you get it working! :)