Well, as usual there were several problems and some misunderstanding.
First of all, thank you for clarifying the naming convention, this clear up a lot of the confusion I was experiencing. I will definitively look at using addIonsRand for future reference. I was initially just interested in getting a usable ion PDB file though I will likely have to add ions again.
“Shortish" version
It seems the issue I was experiencing with leap writing and reading a PDB was correction of terminal groups and unrelated to CL/Cl-. So, using addIons, producing a PDB file and reading this into leap again was not a problem.
As for the packmol problems, packmol did not have a problem building the system. I was having trouble reading the generated PDB file into leap. You are touching on one issue:
Notice, there is a missing blank space between keyword "HETATM" and atom index "1" in your PDB, it must be "HETATM 1" and not "HETATM 1" according to the PDB file format…
This was one thing I failed to correct. This was made worse due to some reckless “greping” messing up the spacing further so that there were two missing spaces, causing enough problems to be noticeable. Not packing with packmol but when reading with leap.
Finally, using a single ion caused a numbering problem with packmol:
ATOM 127 HN2 NHE 8 46.279 49.603 40.091 1.00 0.00
TER
ATOM 128 Cl- Cl- A 1 55.087 55.658 63.028 1.00 0.00
TER
ATOM 129 O WAT B 1 0.000 68.136 31.968 1.00 0.00
Going from one Cl- to first WAT (both being numbered as 1) was not appreciated by leap. Putting my incorrectly aligned chloride after solvent cased issues as the CL/Cl- part bumped up against the atom numbering, casing a reading error. Now, including sequential renumbering of residues “resnumbers 3” in the packmol input, combined with a correctly aligned chloride PDB file, all issues seems to have been resolved.
Thank you very much for you feedback and input.
Best regards
// Gustaf
On 21 Oct 2020, at 19:22, Matias Machado <mmachado.pasteur.edu.uy<mailto:mmachado.pasteur.edu.uy>> wrote:
Dear Gustaf,
I can't understand your issues...
I'm able to place CL ions in a box with Packmol (18.169) using your PDB, which contains 2-letter names:
-----
echo "HETATM 1 CL CL 2 5.829 9.056 -1.558 0.00 0.00 CL" > CL.pdb
cat << EOF > CL_box.pkm
tolerance 3.0
output CL_box.pdb
filetype pdb
avoid_overlap yes
structure CL.pdb
number 20
inside cube 0. 0. 0. 40.
add_amber_ter
end structure
EOF
packmol < CL_box.pkm
-----
Notice, there is a missing blank space between keyword "HETATM" and atom index "1" in your PDB, it must be "HETATM 1" and not "HETATM 1" according to the PDB file format...
Then, I'm able to load the CL box with leap by executing:
-----
cat << EOF > CL_box.leap
source leaprc.water.tip3p
sys = loadpdb CL_box.pdb
saveAmberParmNetcdf sys CL_box.prmtop CL_box.ncrst
quit
EOF
tleap -f CL_box.leap
-----
You seem to have a confusion about names and commands...
Because there is no systematic naming convention regarding ions, you may find different names referring to the same residue in different force field (AMBER, GROMOS, CHARMM...). For example, Cl- CL CL- CLO, they all represent a chloride ion of charge -1.
The way AMBER handle this is by having copies of all topologies... so if you look at file "atomic_ions.lib", which is load by "leaprc.water.tip3p", then you'll find "CL" and "Cl-" residues being the same but for their names.
So this statement "I just noticed that while adding “CL” neutralised the charge just fine, however there are seemingly two “default” options, CL and Cl-." is wrong.
What "addions" command does is adding units to a system while being aware of the charge... it isn't aware of the meaning of the unit, so they can be anything... indeed I use it to and coarse-grained ions to a system!... Hence, these two command render "the same result" in practical terms:
addions CL 0
addions Cl- 0
However, as I mention in several threads, I don't recommend using this "addions" but "addIonsRand".
Best,
Matías
----- Mensaje original -----
De: "Gustaf Olsson" <gustaf.olsson.lnu.se<mailto:gustaf.olsson.lnu.se>>
Para: "AMBER Mailing List" <amber.ambermd.org<mailto:amber.ambermd.org>>
Enviados: Miércoles, 21 de Octubre 2020 9:26:56
Asunto: Re: [AMBER] Adding Ions
I just noticed that while adding “CL” neutralised the charge just fine, however there are seemingly two “default” options, CL and Cl-.
Looking at these options it seems that CL would be “atom” and Cl- “anion”. However, the only example i found for addIons with chloride in the manual used “CL”, making it less likely that this would be the "bonded” state, assuming the “atom” state would be unreasonable. Reusing this option also neutralised the charge.
Am I safe to assume that using “Cl-" would be more appropriate choice with “addIons” (is there any difference)? In that case the remaining issues with reading back a PDB file stems from “packmol” residue numbering and all “problems” have been solved.
Looking at the 3.6 Ions section, is there any reason to source modifications for use with OPC (or TIP3P/ SPC-E for that matter).
For leaper.water.opc:
Monovalent ion parameters for Ewald and TIP4P/EW water from Joung & Cheatham JPCB (2008)
Loading parameters: /Users/guolaa/SOFTWARE/amber20/dat/leap/parm/frcmod.ions234lm_126_tip4pew
Reading force field modification type file (frcmod)
Reading title:
Li/Merz ion parameters of divalent to tetravalent ions for TIP4P/EW water model (12-6 normal usage set)
Though according to the manual:
Two sets of 12-6 LJ parameters for OPC water model (the 12-6 IOD set and the 12-6 HFE set) for 3 monovalent ions (Na+, K+, Cl-) have been developed…
These are not loaded, instead what is described in the manual:
Based on our limited experience, it appears that the Joung/Cheatham ion parameters for TIP4P-EW may be acceptable
Can one safely assume that the parameters loaded with leaper.water.opc/spce/tip3p are reasonable with regard to Cl-?
Best regards
// Gustaf
On 21 Oct 2020, at 09:04, Gustaf Olsson <gustaf.olsson.lnu.se<mailto:gustaf.olsson.lnu.se>> wrote:
Hello Bill, nice to hear from you.
No-go, this might be a problem with the PDB format. Packmol can handle CL just fine, that is not a problem though even if I pack the system with CLX (for example) and then exchange CLX -> CL, the read-frame seems to be is messed up in leap and I end up with errors.
I just used the addIons command instead though this has two effects which could negatively affect the outcome. First, it replaces a solvent molecule and second, you may want to be able to “place” the ion at particular coordinates in the system. Neither of these are of any importance to me right now though I suspect that there is a workaround to be able to use packmol (or something else) to place an ion at a particular coordinate and in this way not removing anything.
All the best
// Gustaf
On 20 Oct 2020, at 15:48, Bill Ross <ross.cgl.ucsf.edu<mailto:ross.cgl.ucsf.edu>> wrote:
Workaround: why not substitute a 3-letter name, packmol it, then
resubstitute?
Bill
On 10/20/20 5:56 AM, Gustaf Olsson wrote:
Hello mailing list
I am trying to construct a system which needs to include a chloride anion to neutralise positive charge.
Doing this in leap with the addIons command works and I am left with a topology and input coordinate file which according to “parmed" has the correct charge (0) and everything looks OK.
Now, I would like to be able to place the ion in the system using packmol. This is where I run into problems.
Writing out a PDB file from leap or parmed produces this
HETATM 1 CL CL 2 5.829 9.056 -1.558 0.00 0.00 CL
This is not usable when trying to build the system with packmol as there is no way to use a two letter abbreviation in a PDB file and read it correctly back into leap, causing all kinds bad.
Is there a reason why I should NOT place the ion in the system using packmol? If not, how can I create a PDB file for a singe Cl- ion, place it in the system using packmol and have a readable file in leap?
Should I do something similar to “WAT = SPC”, as in naming the chloride ion CLX (for example) and setting CLX=CL?
Any help appreciated
Best regards
// Gustaf
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org>
http://lists.ambermd.org/mailman/listinfo/amber
--
Phobrain.com<
http://Phobrain.com>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:AMBER.ambermd.org>
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org<mailto:
AMBER.ambermd.org>
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 22 2020 - 01:30:03 PDT
