Re: Issues with vdW for non existing atom

From: JC Gumbart (gumbart_at_physics.gatech.edu)
Date: Fri Apr 17 2020 - 18:07:46 CDT

Sorry, that I cannot say. I’ve never run QM/MM in NAMD myself. Like with any such error, I would just keep trying to narrow down until getting the simplest system that still triggers it (even nonsensical ones, just for testing).

Best,
JC

> On Apr 17, 2020, at 5:01 AM, Francesco Pietra <chiendarret_at_gmail.com> wrote:
>
> Hi James
>
> Do you find any connection of these MD problems with the scattered problems that I encountered with QMMM? I.e., cases where the params were recognized by MD, not by QMMM. My last mail to Marcelo was not answered yet, hopefully not because of covid19 . It is displayed below (initially it seemed that QMMM was unable to recognize atom types that have a underscore character. But, as told below, later it occurred to me also when no underscore is present.
>
>
> Hi Marcelo, that of the underscore is not the whole story, or it is a wrong story.
>
> Now, with the same receptor and another ligand, which only differs from the workable ligands by a substituent atom at a ring, QMMM immediately crashes with error
>
> Info: 0 NBTHOLE_PAIRS
> FATAL ERROR: UNABLE TO FIND BOND PARAMETERS FOR CG3RC1 NG2S1 (ATOMS 3961 3971)
>
> while those parameters are present in the .str, see last line (I also moved that line as the top at no avail):
>
> read param card
>
> BONDS
> CG3RC1 NG2S1 320.00 1.4340
> CG2R51 NG2R51 400.00 1.3800
> CG2O1 OG2D1 620.00 1.2300
> CG3RC1 NG2R51 220.00 1.4580
> CG3RC1 NG2R53 370.00 1.4500
> NG2S1 HGP1 440.00 0.9970
> CG331 NG2R53 320.00 1.4300
> CG2R53 NG2R53 460.00 1.3800
> CG2R51 CG2R51 410.00 1.3600
> CG3C52 CG3RC1 222.50 1.5240
> CG2R51 CLGR1 350.00 1.7400
> CG2R51 HGR51 350.00 1.0800
> CG2O1 CG2R51 280.00 1.4750
> CG2O1 NG2S1 370.00 1.3450
> CG3RC1 CG3RC1 222.50 1.5230
> CG331 HGA3 322.00 1.1110
> OG311 HGP1 545.00 0.9600
> NG2R53 HGP1 470.00 1.0150
> CG3C52 HGA2 307.00 1.1000
> CG3RC1 HGA1 309.00 1.1110
>
> CG3RC1 OG311 428.00 1.4200
> CG2R53 OG2D1 570.00 1.2350
> CG3RC1 NG2S1 320.00 1.4340
>
> No such problems with minimization and MD.
>
> If it is a misleading error signal, I can't understand its origin.
> cheers
>
> Thanks a lot for you advice
> francesco
>
>
> On Fri, Apr 17, 2020 at 6:15 AM JC Gumbart <gumbart_at_physics.gatech.edu <mailto:gumbart_at_physics.gatech.edu>> wrote:
> It’s in the NBFIX lines of the CGenFF parameter file. I once tried looking into the NAMD code to figure out how to avoid this problem, but the logic was a little more complex than I expected and I gave up. You can either include the related file(s), comment the line out, or make a separate file with just the vdW for any NBFIXed atom types.
>
> Best,
> JC
>
>> On Apr 15, 2020, at 4:55 PM, Francesco Pietra <chiendarret_at_gmail.com <mailto:chiendarret_at_gmail.com>> wrote:
>>
>> Hi Mariano
>>
>> Rerun by adding
>> toppar_all36_carb_glycolipid.str
>> toppar_all36_carb_glycopeptide.str
>>
>> and, to the ligand vdW in the .str:
>>
>> OC2D1 0.00 0.000 0.000
>>
>> it asked for OC2D2
>>
>> OC2D2 0.00 0.000 0.000
>>
>> it asked for
>> OC2D3
>>
>> OC2D3 0.00 0.000 0.000
>>
>> it asked for OC2D4
>>
>> OC2D4 0.00 0.000 0.000
>>
>> now minimization went to completion regularly.
>>
>> However, none of these atom types were accounted for, see the namd log excerpt
>>
>> PREVIOUS VALUES sigma=3.65268 epsilon=0.078 sigma14=3.38542 epsilon14=0.01
>> USING VALUES sigma=3.65268 epsilon=0.078 sigma14=3.65268 epsilon14=0.078
>> Info: SKIPPING rtf SECTION IN STREAM FILE
>> Info: SKIPPING rtf SECTION IN STREAM FILE
>> Info: SUMMARY OF PARAMETERS:
>> Info: 968 BONDS
>> Info: 2885 ANGLES
>> Info: 6702 DIHEDRAL
>> Info: 223 IMPROPER
>>
>> the structure after minimization is OK.
>> There is a iodine atom substituent at sp2 carbon in the ligand and I had to manipulate the .str, removing what concerns lone pair and therefore adjusting the charges. perhaps that has confused namd, but it is unclear. Unclear also whether the toppar for glyco are really needed, but I did not check further.
>>
>> I'll place the ligand with the QM atoms for QMMM, so that I was not much interested in its .str
>>
>> I have no good mind for making a minimal set
>>
>> regards
>> francesco
>>
>>
>> On Wed, Apr 15, 2020 at 5:54 PM Mariano Spivak <marspivak_at_gmail.com <mailto:marspivak_at_gmail.com>> wrote:
>> Hello Francesco,
>>
>> If you would like to share a minimal set of files to reproduce this error, I can take a look and help you.
>>
>> Best regards
>>
>> Mariano Spivak, Ph.D.
>> Theoretical and Computational Biophysics Group
>>
>> Beckman Institute, University of Illinois.
>>
>> mariano_at_ks.uiuc.edu <mailto:mariano_at_ks.uiuc.edu>
>> mspivak_at_illinois.edu <mailto:mspivak_at_illinois.edu>
>>
>> On Wed, Apr 15, 2020 at 10:34 AM Francesco Pietra <chiendarret_at_gmail.com <mailto:chiendarret_at_gmail.com>> wrote:
>> Hello
>> As you can see from the log file below, namd 2.13, on minimization, complains about the lack of vdW for atom type OC2D1. Actually such atom type, if it exists at all, is not present in either the .str file for my ligand or the general .psf file for my receptor made of nuclei/prot, on attempt to minimization, generated with CHARMM36 FF jul19, i.e the latest, which includes CGenFF 4.3.
>>
>> As I plan to carry out QMMM, I added the to the .str vdW the line
>> OC2D1 0.00 0.000 0.000
>> getting the same request of vdW for atom type OC2D1
>>
>> Warning: DUPLICATE vdW ENTRY FOR IGR1
>> PREVIOUS VALUES sigma=3.99123 epsilon=0.52 sigma14=3.99123 epsilon14=0.52
>> USING VALUES sigma=3.52796 epsilon=0.23 sigma14=3.52796 epsilon14=0.23
>> Warning: DUPLICATE vdW ENTRY FOR CG3RC1
>> PREVIOUS VALUES sigma=3.56359 epsilon=0.032 sigma14=3.38542 epsilon14=0.01
>> USING VALUES sigma=3.56359 epsilon=0.032 sigma14=3.56359 epsilon14=0.032
>> Warning: DUPLICATE vdW ENTRY FOR NG2S1
>> PREVIOUS VALUES sigma=3.29633 epsilon=0.2 sigma14=2.76179 epsilon14=0.2
>> USING VALUES sigma=3.29633 epsilon=0.2 sigma14=3.29633 epsilon14=0.2
>> Warning: DUPLICATE vdW ENTRY FOR CG3C52
>> PREVIOUS VALUES sigma=3.59923 epsilon=0.06 sigma14=3.38542 epsilon14=0.01
>> USING VALUES sigma=3.59923 epsilon=0.06 sigma14=3.59923 epsilon14=0.06
>> Warning: DUPLICATE vdW ENTRY FOR OG2D1
>> PREVIOUS VALUES sigma=3.02906 epsilon=0.12 sigma14=2.49452 epsilon14=0.12
>> USING VALUES sigma=3.02906 epsilon=0.12 sigma14=3.02906 epsilon14=0.12
>> Warning: DUPLICATE vdW ENTRY FOR CG331
>> PREVIOUS VALUES sigma=3.65268 epsilon=0.078 sigma14=3.38542 epsilon14=0.01
>> USING VALUES sigma=3.65268 epsilon=0.078 sigma14=3.65268 epsilon14=0.078
>> FATAL ERROR: DIDN'T FIND vdW PARAMETER FOR ATOM TYPE OC2D1
>>
>> thanks for advice
>> francesco pietra
>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:13 CST