[gmx-developers] Re: g_dist and vsite

[Anton Feenstra]

Ilya Chorny wrote:
> I have been having problems calculating the distance between CA and CB 
> atoms in ALA using g_dist. The trajectory was run using vsite so the 
> mass of CB in ALA is set to 0. Since g_dist calculates distances between 
> the center of mass of the groups if one of the groups has a mass of zero 
> there will be an error.
> 
> I am going to edit gmx_dist as follows. Does anyone see any glaring 
> issues with doing this?

You're assuming a mass of one whenever the mass is 0, I'm not sure that 
makes much sense. If for example the rest of your system is mainly 
carbon, your v-sites will not contribute to the COM as much as you may 
expect. (you don't have that in your analysis, but others may)

I can think of two better (IMHO) solutions:
- use the center of geometry instead - no masses, no problem. You may 
make this an option in gmx_dist.
- use the 'physical' masses instead. This could be done independently of 
the analysis tool, by constructing an equivalent topology without 
v-sites. There are some routines in the libraries (not sure now exactly 
where), that actually do a lookup of atom mass based on the atom names. 
This is not fool-proof in the general case, but works good enough for 
e.g. proteins, or other systems with systematic and/or predictable atom 
names.


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | IBIVU/Bioinformatics - Free University  Amsterdam     |
|(   |   )| | | De Boelelaan 1083A - 1081 HV Amsterdam - Netherlands  |
| \_/ \_/ | | | Tel +31 20 59 87783 - Fax +31 20 59 87653 - Room P136 |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "It's Because Of the Metric System" (Pulp Fiction)    |
|_____________|_______________________________________________________|