Support for cgenff 4.0 halogen groups (building and simulation)

[ismaelresp]

Last releases of CHARMM force-field now work with CGenFF 4.0. These provide lone pairs for halogen atoms and paramchem 2.0 already provides support for producing analogy based parameters for CGenFF 4.0.
Are there any plans on supporting long pairs in ACEMD and/or HTMD?

[stefdoerr]

You mean as in parameterize parameterizing molecules with CGenFF?
Otherwise if you pass the right parameters to charmm.build it should build it fine assuming psfgen has not changed dramatically in the last versions and ACEMD simulate it fine. Do you have some example?

[ismaelresp]

I was asking only for building and running simulations, not about the parametrize tool.
I do not have any examples at the moment as I was unsure even if psfgen was going to work correctly with them. In any case, PSFs can be generated with CHARMM software.

[stefdoerr]

I'm sorry I don't have any experience with that then. You will have to give it a try :)

[ismaelresp]

Anyway, I found in CHARMM documentation on PSF lone pair syntax and how lone pairs should behave:
https://www.charmm.org/charmm/documentation/by-version/c42b1/params/doc/lonepair/

Psfgen seems to at least add the lone pair dummy atoms but does not seem to locate them correctly and gives the error
ERROR! FAILED TO RECOGNIZE LONEPAIR. Line 107: LONEPAIR COLI LP1 Cl1 C9 DIST 1.6400 SCAL 0.0

[ismaelresp]

Here is the paper for CGENFF lone pair addition: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5053860/

[ismaelresp]

NAMD gives support for them in its last version:
https://www.ks.uiuc.edu/Research/namd/2.13/features.html

[ismaelresp]

Anyway, I found in CHARMM documentation on PSF lone pair syntax and how lone pairs should behave:
https://www.charmm.org/charmm/documentation/by-version/c42b1/params/doc/lonepair/

Psfgen seems to at least add the lone pair dummy atoms but does not seem to locate them correctly and gives the error
ERROR! FAILED TO RECOGNIZE LONEPAIR. Line 107: LONEPAIR COLI LP1 Cl1 C9 DIST 1.6400 SCAL 0.0

I used the psfgen that is shipped with vmd 1.9.3.

[ismaelresp]

Anyway, I found in CHARMM documentation on PSF lone pair syntax and how lone pairs should behave:
https://www.charmm.org/charmm/documentation/by-version/c42b1/params/doc/lonepair/
Psfgen seems to at least add the lone pair dummy atoms but does not seem to locate them correctly and gives the error
ERROR! FAILED TO RECOGNIZE LONEPAIR. Line 107: LONEPAIR COLI LP1 Cl1 C9 DIST 1.6400 SCAL 0.0

I used the psfgen that is shipped with vmd 1.9.3.

Same result with charmm.build

[ismaelresp]

LONEPAIR COLI

Only colinear seem to be needed at the moment.

[giadefa]

Maybe we just need to use the latest psfgen in htmd.

…

On Tue, 19 Feb 2019 at 23:21, Ismael Rodriguez Espigares < ***@***.***> wrote: LONEPAIR COLI Only colinear seem to be needed at the moment. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#853 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AHaqOvpKgkCJBbjoEV4NqE6WqS8JJXvSks5vPHj3gaJpZM4bDNvH> .

-- <http://www.acellera.com> <https://twitter.com/acellera> <https://www.youtube.com/user/acelleracom> <https://www.linkedin.com/company/2133167?trk=tyah&trkInfo=clickedVertical%3Acompany%2CclickedEntityId%3A2133167%2Cidx%3A2-1-2%2CtarId%3A1448018583204%2Ctas%3Aacellera> <https://www.acellera.com/md-simulation-blog-news/> <http://is.gd/1eXkbS>

[ismaelresp]

Maybe we just need to use the latest psfgen in htmd.
…
On Tue, 19 Feb 2019 at 23:21, Ismael Rodriguez Espigares < @.***> wrote: LONEPAIR COLI Only colinear seem to be needed at the moment. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#853 (comment)>, or mute the thread https://github.com/notifications/unsubscribe-auth/AHaqOvpKgkCJBbjoEV4NqE6WqS8JJXvSks5vPHj3gaJpZM4bDNvH .
-- http://www.acellera.com https://twitter.com/acellera https://www.youtube.com/user/acelleracom https://www.linkedin.com/company/2133167?trk=tyah&trkInfo=clickedVertical%3Acompany%2CclickedEntityId%3A2133167%2Cidx%3A2-1-2%2CtarId%3A1448018583204%2Ctas%3Aacellera https://www.acellera.com/md-simulation-blog-news/ http://is.gd/1eXkbS

They have not implemented lonepairs in psfgen yet (cannot parse LONEPAIR statement):
https://charm.cs.illinois.edu/gerrit/gitweb?p=namd.git;a=blob;f=psfgen/src/charmm_parse_topo_defs.c;h=27189247e436b5f5cace1062aa9691e2693b4256;hb=refs/heads/devel

But they support PSF lonepairs in the same format as Drude on NAMD:
https://charm.cs.illinois.edu/gerrit/gitweb?p=namd.git;a=commit;h=846dbb0e27253d4b0e16c4c5c5616734c3d66dc9

[ismaelresp]

Anyway, I found in CHARMM documentation on PSF lone pair syntax and how lone pairs should behave:
https://www.charmm.org/charmm/documentation/by-version/c42b1/params/doc/lonepair/
Psfgen seems to at least add the lone pair dummy atoms but does not seem to locate them correctly and gives the error
ERROR! FAILED TO RECOGNIZE LONEPAIR. Line 107: LONEPAIR COLI LP1 Cl1 C9 DIST 1.6400 SCAL 0.0

I used the psfgen that is shipped with vmd 1.9.3.

Same result with charmm.build

After building a system with CHARMM software and testing it in NAMD, I reach the conclusion that psfgen does not support lone pairs at all as a new field has to be added at the end of the PSF file (Drude format). Otherwise, NAMD does nothing with the lone pair dummy atoms.
Check also here.
So, neither PSFGEN (Git-2019-02-19) nor HTMD support cgenff 4.0 right now for aromatic halogen groups.

[ismaelresp]

I tested NAMD and ACEMD with my system built with CHARMM software. It seems working in NAMD, but not in ACEMD. In ACEMD, I cannot get passed equilibration (it seems that the barostat does not like zero-mass dummy atoms), but I've got the coordinates after minimization and it seems that ACEMD is ignoring the lone pair defined in the PSF as you can see in the two following images (links bellow):
https://drive.google.com/open?id=1hzu07qvJB370WEzMbr_tk9f3ic-xUwY5
https://drive.google.com/open?id=1Py2V02LEjkVDxo8S7p-i2pB2Gn7vjll0

Here you have a link to my testing system and scripts:
https://drive.google.com/open?id=1GGBKI3D8-BuiDQ-BLi_CNXeMq3orJfF8

The testing system only contains clozapine parametrized with paramchem 2.2.0 and cgenff4.0 solvated in TIP3 water and NaCl.

[ismaelresp]

So, to sum up, neither psfgen, htmd nor acemd seem to support cgenff4.0 lone pairs right now. Hence, CHARMM FF versions after February 2016 (not included) cannot be used as they are with paramchem parameters for ligands with halogen groups (and I do not know if any modifications are needed to make them backwards compatible with cgenff3.0.1) with these 3 software packages for the moment.

Any plans for ACEMD compatibility with cgenff4.0 and halogen groups?

[ismaelresp]

Here, some info about how the lone pair PSF section works:

https://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2013-2014/1203.html
https://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l.2018-2019/1082.html

https://charm.cs.illinois.edu/gerrit/gitweb?p=namd.git;a=commit;h=846dbb0e27253d4b0e16c4c5c5616734c3d66dc9
https://charm.cs.illinois.edu/gerrit/gitweb?p=namd.git;a=blobdiff;f=src/structures.h;h=7f636cd3ad9c1980028341be75a30a1f8f45dba0;hp=1bdccbe58c6a3e1d56f75e3679961cffd18c020f;hb=846dbb0e27253d4b0e16c4c5c5616734c3d66dc9;hpb=cffde1d70bcbfe87d18aff6847558d9281d6f206
https://charm.cs.illinois.edu/gerrit/gitweb?p=namd.git;a=blobdiff;f=src/Molecule.C;h=d8d13ead41e470bfa7a9c0e66e83837a9b8f5333;hp=96069c380890dad013314aa31fc98ab173e858d4;hb=846dbb0e27253d4b0e16c4c5c5616734c3d66dc9;hpb=cffde1d70bcbfe87d18aff6847558d9281d6f206

And here there is an example of a PSF lone pair section:

     1         3 !NUMLP NUMLPH !#NUMLP=total_number_of_long_pairs NUMLPH=total_number_of_lone_pair_hosts+NUMLP
     2         1   F   1.64000       0.00000       0.00000    !# number_of_LPH_for_current_LP(2_OR_3) LP_index_in_LP_LPH_atom_index_list weight=F distance/value1 angle/value2 dihedral/value3
    44        23        12 !#(LP LPH1 LPH2 [LPH3]) list (atom indexes). Length=NUMLPH

Lone pair hosts (LPH): atoms whose coordinates are used for determining the lone pair position.