VMD-L Mailing List

From: JC Gumbart (gumbart_at_physics.gatech.edu)
Date: Fri May 18 2018 - 09:51:53 CDT

(copying the list for community benefit)

The problem is that the element for hydrogens wasn’t set correctly for whatever reason. So in your QM calculations, all the hydrogens were marked as atom type X and summarily ignored.

The simplest way around this is to load your PSF/PDB and type

set sel [atomselect top “name \”H.*\””]
$sel set element H
set all [atomselect top all]
$all writepdb new.pdb

And then use that for all the QM setup, analysis, etc.

As a fix, we need to figure out if we want to deal with this within FFTK or PSFGen, the latter of which appears to not add element names to any newly created atoms.

Best,
JC

> On May 17, 2018, at 4:13 PM, <dalpizar_at_fbio.uh.cu> <dalpizar_at_fbio.uh.cu> wrote:
>
> Here I send you the files that I used like second way, I had to make a .zip, this contain yhe .log from the optimization with the Water Interactions, the pdb, the psf, the .str and the file par_all36_cgenff.prm If you need anything else please let me now, I hope this will be useful
>
> Thanks for your time
> Daniel
>
> El 2018-05-17 15:33, JC Gumbart escribió:
>> Whatever is the smallest set that reproduces the error is fine.
>> Best,
>> JC
>>> On May 17, 2018, at 2:41 PM, <dalpizar_at_fbio.uh.cu> <dalpizar_at_fbio.uh.cu> wrote:
>>> Well, I made many probes whit the aim of resolve this issue and in the process I generated many files, I optimized the Water interactions in many ways:
>>> The .psf and .pdb that I use were generated with psfgen script in vmd.
>>> i) I use the button Guees to predict the Donor and Aceptor indices in the molecule, optimize this and made the Charges optimization whit this files
>>> ii) I put the donor and aceptor indices that in my .str, obtained in CGenFF, shown high penalties because those are the charges that I am interested in opimize, I optimize this Water Interactions and made the Charges optimization whit this files
>>> iii) I put all atoms from my molecule in the donor and aceptor indices and optimize all this interactions
>>> In each of this ways in the Charge Opt I made the same:
>>> i) input the .str obtained in CGenFF with the penalties and the parameters file par_all36_cgenff.prm from CHarmmFF
>>> ii) In charges Constrain section I only put the atoms that I want optimize and fix the low and high bounds when I put calculate from psf I obtain Net Charges 0
>>> iii) In QM Target Data Sections I input the files that corrrespond in each box obtained in each ways that I mentioned before for the Water interaction section
>>> At last I put run Optimization and in all case the error is the same:
>>>>> vecscale: parameters must have data
>>>>> vecscale: parameters must have data
>>>>> while executing
>>>>> "vecscale 1E-10 [lindex $dipoleQMcoords $ind]"
>>>>> (procedure "::ForceFieldToolKit::ChargeOpt::optCharges" line 98)
>>>>> invoked from within
>>>>> "$function $p0list"
>>>>> (procedure "construct_initial_complex" line 13)
>>>>> invoked from within
>>>>> "construct_initial_complex [lindex $args 0]"
>>>>> (procedure "::Optimize::Opt0::handle" line 89)
>>>>> invoked from within
>>>>> "$opt initsimplex [lrange $chargeInit 0 end-1] "
>>>>> (procedure "::ForceFieldToolKit::ChargeOpt::optimize" line 410)
>>>>> invoked from within
>>>>> "::ForceFieldToolKit::ChargeOpt::optimize"
>>>>> (procedure "::ForceFieldToolKit::gui::coptRunOpt" line 55)
>>>>> invoked from within
>>>>> "::ForceFieldToolKit::gui::coptRunOpt "
>>>>> invoked from within
>>>>> ".fftk_gui.hlf.nb.chargeopt.runOpt invoke "
>>>>> invoked from within
>>>>> ".fftk_gui.hlf.nb.chargeopt.runOpt instate {pressed !disabled} { .fftk_gui.hlf.nb.chargeopt.runOpt state !pressed; .fftk_gui.hlf.nb.chargeopt.runOpt in..."
>>>>> (command bound to event)
>>> So, what files I send to you? all of them or only an especific set from the before mentioned? Do you see any mistake when I input the files?
>>> Regards
>>> Daniel
>>> El 2018-05-16 23:50, JC Gumbart escribió:
>>>> Can you make a tgz of the files and send it to me here? If it’s not
>>>> too large that is. And then tell me exactly what you did.
>>>> Best,
>>>> JC
>>>>> On May 16, 2018, at 2:55 PM, <dalpizar_at_fbio.uh.cu> <dalpizar_at_fbio.uh.cu> wrote:
>>>>> Hi JC:
>>>>> I made the Charge Opt whit the vmd version 1.9.4 alpha builds as you sugested and I obtained the same results Tell me what do you need and where I can send you the files The error is next:
>>>>> vecscale: parameters must have data
>>>>> vecscale: parameters must have data
>>>>> while executing
>>>>> "vecscale 1E-10 [lindex $dipoleQMcoords $ind]"
>>>>> (procedure "::ForceFieldToolKit::ChargeOpt::optCharges" line 98)
>>>>> invoked from within
>>>>> "$function $p0list"
>>>>> (procedure "construct_initial_complex" line 13)
>>>>> invoked from within
>>>>> "construct_initial_complex [lindex $args 0]"
>>>>> (procedure "::Optimize::Opt0::handle" line 89)
>>>>> invoked from within
>>>>> "$opt initsimplex [lrange $chargeInit 0 end-1] "
>>>>> (procedure "::ForceFieldToolKit::ChargeOpt::optimize" line 410)
>>>>> invoked from within
>>>>> "::ForceFieldToolKit::ChargeOpt::optimize"
>>>>> (procedure "::ForceFieldToolKit::gui::coptRunOpt" line 55)
>>>>> invoked from within
>>>>> "::ForceFieldToolKit::gui::coptRunOpt "
>>>>> invoked from within
>>>>> ".fftk_gui.hlf.nb.chargeopt.runOpt invoke "
>>>>> invoked from within
>>>>> ".fftk_gui.hlf.nb.chargeopt.runOpt instate {pressed !disabled} { .fftk_gui.hlf.nb.chargeopt.runOpt state !pressed; .fftk_gui.hlf.nb.chargeopt.runOpt in..."
>>>>> (command bound to event)
>>>>> Thanks
>>>>> Daniel
>>>>> El 2018-05-11 18:56, JC Gumbart escribió:
>>>>>> Your description is too vague to be of much use. Can you first try
>>>>>> obtaining the latest version of VMD (one of the 1.9.4 alpha builds)
>>>>>> and trying again there? If that doesn’t work, probably the easiest
>>>>>> way is to just send me the files off-list with instructions on how to
>>>>>> reproduce your error.
>>>>>> Best,
>>>>>> JC
>>>>>>> On May 11, 2018, at 4:17 PM, dalpizar_at_fbio.uh.cu wrote:
>>>>>>> Hello Im trying to optimize charges in fftk using VMD 1.9.3, but when I put Run optimization the program give me back this issue
>>>>>>> Error: vecscale: parameters must have data
>>>>>>> Whats this mean? and What I can do to resolve this problem?
>>>>>>> Thanks
>>>>>>> Daniel
> <Files.zip>