[gmx-users] Bug in g_mindist 3.3.1

[Volker Knecht]

Dear David, dear all,

>Volker Knecht wrote:
>  
>
>>there seems to be a bug in g_mindist version 3.3.1. When using g_mindist 
>>to determine the minimal distance between two molecules in solution as a 
>>function of time, the output fluctuates around 0.1nm although the 
>>molecules are separated for part of the time. When using version 3.2.1 
>>instead, the output is fine (separation on the nm range for part of the 
>>time in agreement with visual inspection).
>>
>>Unfortunately, g_mindist version 3.2.1 gives the right output if it 
>>runs, but in some cases crashes with a segmentation fault as reported in 
>>the mailing list previously.  Therefore it would very helpful if the 
>>apparent bug in version 3.3.1 could be fixed.
>>
>>Thanks in advance.
>>
>>Kind regards,
>>Volker
>>
>>    
>>
David van der Spoel wrote:

>Are you certain it is not due to periodic boundary conditions, i.e. te 
>distance is reallly correct if PBC are taken into account?
>  
>
Yes, I'm quite sure. When I take a snapshot and multiply it in all 
directions using genconf and look at the resulting pdb using rasmol, the 
two molecules are clearly separated. This is also supported by the 
results from g_mindist version 3.2.1 which however suffers from 
different problems as stated above. g_mindist version 3.3.1 yields a 
minimum distance of 0.1nm for this snapshot.

Volker

-- 
Dr. Volker Knecht
Theory Department
Max Planck Institute of Colloids and Interfaces
Research Campus Golm
14424 Potsdam, Germany

Phone: +49-331-567-9610
Fax: +49-331-567-9612

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060626/ede5bbbe/attachment.html>