Page 1 of 1
CHARMM FF LONEPAIR
Posted: Fri Apr 26, 2019 7:02 am
by khaos
Dear OpenMM developers,
It seems like that current OpenMM only supports 'relative' and 'bisector' types of LONEPAIR for Drude FF.
Do you have any plan to implement the support of 'COLINEAR' type LONEPAIR for CGenFF?
Best,
Jumin
Re: CHARMM FF LONEPAIR
Posted: Fri Apr 26, 2019 10:19 am
by peastman
The LocalCoordinatesSite class is very flexible. It should be able to support any kind of lone pair used in CHARMM and many others. See
http://docs.openmm.org/latest/userguide ... tual-sites.
Or are you talking about the implementation of the Drude force field in charmm_polar_2013.xml? If so, are you saying there are residues that are missing lone pairs? Which ones?
Re: CHARMM FF LONEPAIR
Posted: Mon Apr 29, 2019 5:32 am
by khaos
Sorry for confusing you.
No, I am talking about CHARMM PSF parser.
In wrappers/python/simtk/openmm/app/charmmpsffile.py line 374:
https://github.com/pandegroup/openmm/bl ... le.py#L374
Code: Select all
if numlp != 0 or numlph != 0:
lp_distance_list=[]
lp_angle_list=[]
lp_dihe_list=[]
for i in range(numlp):
lpline = holder[i].split()
if len(lpline)!=6 or lpline[0] != '3' or lpline[2] != 'F' or int(lpline[1]) != 4*i+1 :
raise CharmmPSFError('Lonepair format error')
else:
lp_distance_list.append(float(lpline[3]))
lp_angle_list.append(float(lpline[4]))
lp_dihe_list.append(float(lpline[5]))
COLINEAR type LONEPAIR only has 2 hosts so that lpline[0] is 2, and it does not have angle and dihedral terms.
Re: CHARMM FF LONEPAIR
Posted: Mon Apr 29, 2019 9:13 am
by peastman
I'm not familiar with that code. It was written by Jing Huang:
https://github.com/pandegroup/openmm/pull/2120. Perhaps contact him and ask him about adding the other type?
Re: CHARMM FF LONEPAIR
Posted: Mon Apr 29, 2019 9:37 am
by khaos
OK. Thank you.
I will contact him.