Dear msi2lmp tool users,
I’m conducting a simulation involving 3,3’-Diaminodiphenyl Sulfone (33dds). I build this molecule in Material studio (version 8.0), apply pcff force field and get the (.car) and (.mdf) files (see the attachments). But when I use msi2lmp tool to convert these MS files to the lammps data file, the parameters of some bonds is missing (i.e., being set as 0.0) in the lammps data file (see the attachment and screen snapshot). The command I used for the conversion is: msi2lmp 33dds -c 2 -f pcff -i
I checked the pcff force field file in MS. The parameters of these missing bonds is not explicitly specified and they should be automatically generated based on ad hoc model in MS. It seems that these missing bonds cannot be converted by the msi2lmp tool. Is there any way to obtain the parameters for these missing bonds?
PS: I also try cvff force field, the same problem occurs. But if the molecule is simple, such as a benzene molecule, it works well.
Any suggestion or comment is appreciated.
Regards,
----Tengfei LIANG
33dds.car (2.43 KB)
33dds.mdf (2.86 KB)
33dds.data (15.7 KB)
Dear msi2lmp tool users,
I’m conducting a simulation involving 3,3'-Diaminodiphenyl Sulfone
(33dds). I build this molecule in Material studio (version 8.0), apply pcff
force field and get the (.car) and (.mdf) files (see the attachments). But
when I use msi2lmp tool to convert these MS files to the lammps data file,
the parameters of some bonds is missing (i.e., being set as 0.0) in the
lammps data file (see the attachment and screen snapshot). The command I
used for the conversion is: msi2lmp 33dds -c 2 -f pcff -i
I checked the pcff force field file in MS. The parameters of these missing
bonds is not explicitly specified and they should be automatically
generated based on ad hoc model in MS. It seems that these missing bonds
cannot be converted by the msi2lmp tool. Is there any way to obtain the
parameters for these missing bonds?
are you talking about "equivalences" or "auto_equivalences"? msi2lmp has
support for the former but not the latter.
please also note, that the force field parameter files in MS are
proprietary, licensed material. msi2lmp only contains settings and
parameters from the published literature, so you may want to try with the
.frc file bundled with MS instead.
msi2lmp is at its core software that was written over 10 years ago and has
primarily seen updates to remove bugs, improve performance and make it
compliant with changes in LAMMPS. better compatibility would require
programming and validation, which requires access to MS.
axel.
This sounds similar to several other recent posts on the mailing list regarding msi2lmp and EMC (“insight2lammps.pl”):
https://sourceforge.net/p/lammps/mailman/message/35698640/
https://sourceforge.net/p/lammps/mailman/message/29578563/
https://sourceforge.net/p/lammps/mailman/message/35700652/
Assuming this is, indeed, the same as the error you are having, then the solution (according to Marcus Martin) is to use the “bond” (auto) equivalences to lookup the missing bond lengths (in bond-bond and bond-angle, end_bond-torsion, middle_bond-torsion, bond-bond_1_3 interactions).
Similarly I am guessing you must use “angle” (auto) equivalences to look up missing bond angles (in angle-torsion and angle-angle-torsion, and angle-angle interactions).
It seems as though like msi2lmp get confused whenever the (auto) equivalences used for looking up bonds differs from the equivalences used for looking up angles (or dihedrals or impropers).
So, until these tools get fixed, you will have to look them up yourself manually by reading the .FRC file carefully.
I am trying to get these force-fields added to moltemplate so that nobody has to use materials studio or msi2lmp. Moltemplate had this same bug as well, and I patched it today. Unfortunately moltemate is crashing for other reasons. I hope to have moltemplate working and tested examples with these MSI force fields within a couple months (depending on when I can find time off to work on it).
Andrew
I apologize if the error I am describing is not the same as the one you are reporting.