Topic Options
#36570 - 08/18/17 08:45 AM patch residue problem
madape Offline
Forum Member

Registered: 08/13/17
Posts: 3
Dear all,

I got a simple but tricky question: Simulation metalloprotein.

After been reading in the forum possibilities to ensure zinc coordination:

1. Rely on ELEC and VDW, indeed stone parameters for zinc ion. This option does not work properly for me, some coordination bonds are lost.

2. I launch QM/MM calculations, but its too expensive.

Therefore I decided to choose two options:

Add NOE restraints or Patch residue.

My question comes from this last because I am not sure how to implemented and I got an error. My files look like:

* Append the PDBs
*

! Read topology and parameter files
stream toppar.str

! Read PSF and coordinates from file
read psf card name step1_pdbreader.psf
read coor card name step1_pdbreader.crd

patch ZNSP HETA 62 PROA 50 PROA 57 PROA 59 PROA 60 setup warn
autogen angle dihe

coor init sele hydrogen end
hbuild

write psf card name final.psf
write coor card name final.crd
stop


and my output:


Warning from PSFSUM: The sum of charges ( -4.280000) is not an integer


***** LEVEL 0 WARNING FROM *****
***** Total charge not an integer
******************************************
BOMLEV ( 0) IS REACHED - TERMINATING. WRNLEV IS 5


Please, can anyone help me? Because I cannot keep no it
Thank you in advance


Attachments
toppar_all22_cys.txt (363 downloads)


Top
#36571 - 08/18/17 11:05 AM Re: patch residue problem [Re: madape]
rmv Online   content

Forum Member

Registered: 09/17/03
Posts: 8097
Loc: 39 03 48 N, 77 06 54 W
The warning means that applying the patch results in a non-integer charge for the molecule; you need to take the charges of the deleted hydrogens into account.
_________________________
Rick Venable
computational chemist


Top
#36574 - 08/30/17 02:09 AM Re: patch residue problem [Re: madape]
madape Offline
Forum Member

Registered: 08/13/17
Posts: 3
Okay, thank you Rick.

Top
#36598 - 09/24/17 02:49 AM Re: patch residue problem [Re: madape]
madape Offline
Forum Member

Registered: 08/13/17
Posts: 3
Hello Rick,

I tried to skip it by using BOMLEV -2. However, I am not sure if I can get in troubles with. How could I solve the integer charges? How to take into account those delete H charges?

Thank you

Top
#36600 - 09/24/17 11:16 AM Re: patch residue problem [Re: madape]
rmv Online   content

Forum Member

Registered: 09/17/03
Posts: 8097
Loc: 39 03 48 N, 77 06 54 W
Using a more negative bomlev is not a solution, it is an invitation to further trouble.

It looks like you've taken some residues and patches from different sources and tried to combine them, which may be unwise. The patch assumes the CYS residues with the H still attached, while the residues appear to be CYS modified to be thiolate, which is not consistent.

I think you need to slow down and evaluate this issue much more carefully, and learn more about how the CHARMM force field parameters are derived. You may wish to contact people who've published CHARMM simulations involving Zn coordination via patching.

If this will be an important part of your work, you really need to take the time and effort to do the best job you can do.
_________________________
Rick Venable
computational chemist


Top

Moderator:  alex, chmgr, John Legato