Re: Tcl, minimization and MD (Warning: I am a lazy ignorant)

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Wed Jun 21 2006 - 12:34:40 CDT

Hi Miro,
I must apologize, I think I missed the main point of your previous
email. I appreciate you bringing this odd behavior with prolines to my
attention; I will take a look and see if I can reproduce what you've
described. Can you clarify, I'm sorry, what your obvious question is
that you're having trouble finding the answer to?
Peter

moman_at_titus.u-strasbg.fr wrote:
> Thanks a lot to Mark, JC and Pedro for given me the right (and expected)
> anwser.
>
> Pedro: I have gone once more (quickily, I must confess) throgh the NAMD
> tutorial (namely the las chapter as you suggested), but I could not find
> an explicit anwser to my (obvious) question:
>
> http://www.ks.uiuc.edu/Training/Tutorials/namd/namd-tutorial-html/
>
> Surely it is somewhere there, but sitll I can no find it. Never mind.
> Thanks a million.
>
> Thaks to Peter as well for always being so regardful and helpful. However,
> you got everything this time wrong (mea culpa!). This was not a troll. I
> just allowed myself to use a little bit of irony (about last week's
> friendly controversy) WHILE asking a honest question AND making you note
> what it might be a little bug in the VDM autopsfgen (remarkable) feature:
> it seems it systematically asigns wrong coordinates to proline hydrogens
> (it removes them from the original file and add them again). I tried a
> couple of different protein files generated/converted with either
> OpenBabel, SPDBV and MOE and got similar results (wrong geometries) for
> the prolines.
>
> This only shows the point that I wanted to highlight with my irony: we are
> all (kind of) human ;-) and we are all in the same boat, aren't we?
>
> Cheers,
>
> Miro
>
>
>
>> Also... while I did get a chuckle out of the rest of your message, I
>> would recommend not feeding the troll. Even he would agree that it is
>> better for you to come with what you know and an open mind, than to be
>> unduly self deprecating.
>> Peter
>>
>> Edelmiro Moman wrote:
>>
>>> Hello,
>>>
>>> I do not know a work of Tcl scripting (or almost any other scripting
>>> language). In consequence, when I try to use a modified version of the
>>> NAMD sample.conf file, I get all kind of warning messages (if I use
>>> constraints, they tell me about bad geometries; otherwise they
>>> complain about high velocities) and the simulation is immediately
>>> aborted. I (ignorant) suspect that NAMD is starting directly with the
>>> MD simulation without doing a previous minimization (as I would wish).
>>> This is the relevant (I guess) fragment of the file:
>>> ______________________________________________________________
>>>
>>> #############################################################
>>> ## EXECUTION SCRIPT ##
>>> #############################################################
>>>
>>> # Minimization
>>> if {0} {
>>> minimize 100
>>> reinitvels $temperature
>>> }
>>>
>>> run 50000 ;# 100ps
>>> ______________________________________________________________
>>>
>>> First silly question: Am I right (by chance)?
>>>
>>> Second silly question: What can I do?
>>>
>>> Some clues: If the system has been previously minimized the simulation
>>> uses to runs uneventfully; but (!) Autopsfgen removes the hydrogens
>>> from my system and add them again, as a result of which the geometry
>>> of the prolines uses to be completely wrong and that takes me to the
>>> original problem.
>>>
>>> Kind regards,
>>>
>>> Miro
>>>
>>> P.S.: I am not as lazy as I am ignorant. However, as I am aware that
>>> my messages can hurt certain sensibilities, I have include a warning
>>> in the subject so that sensitive people do not have to read them.
>>> About unnecessarily increasing network traffic, I am very sorry about
>>> that and I plan to punish myself as soon as I can think of a suitable
>>> way of doing it.
>>>
>>>
>>>
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:14 CST