VMD-L Mailing List

From: JC Gumbart (gumbart_at_ks.uiuc.edu)
Date: Wed Sep 09 2015 - 10:42:18 CDT

Hi Kevin,

Sorry for not seeing your email earlier - I rarely have time to check the mailing lists anymore. This is indeed a longstanding problem with stride and large structures. The way I have solved it in the past is to use a script to feed individual segments to stride such that their SS is calculated correctly. There is a link for a script on the VMD script page:

http://www.ks.uiuc.edu/Research/vmd/script_library/scripts/stride_workaround <http://www.ks.uiuc.edu/Research/vmd/script_library/scripts/stride_workaround>

But it seems to have disappeared (or never was there?); John might remember. If not, I have one I can provide to him for the website.

Once you have the correct SS in VMD, you can hack the ssrestraints.tcl plugin to read from the current molecule instead of loading a new one.

———

All of that being said, I see in the plugins directory that ssrestraints.tcl calls another set of procs in ssrestraints_stride.tcl that are supposed to be more robust than the normal way of calling stride? I am not familiar with it, so I can’t comment.

Best,
JC

> On Sep 9, 2015, at 12:18 AM, Kevin C Chan <cchan2242-c_at_my.cityu.edu.hk> wrote:
>
> Dear Users,
>
> Sorry for digging this old thread. But I have not yet found the solution for this. Maybe let me explain my question again in case this email cannot go into the right thread:
>
> I have a quite large protein (a protein lattice in fact) and it is from rigid-body docking repeating units into a density map. The docking was very crude and we are going to do MDFF on it. However during the preparation of psf files, I found VMD fails to generate a New Cartoon representation for it which requires guessing secondary structures of it (so of course coloring by sec struc also fails).
>
> Then I narrowed down the problem to those single units of protein and found all of them are okay to be generated such representation individually.
>
> I have examined the proteins and there are some very closely placed atoms that might lead to difficult or wrong guesses.
>
> Interestingly, when I thought it is due to bad initial structure and continued with my MDFF and obtained a much better structure, the final structure with no obvious bad atom positions still fails to generate the secondary structure representation. Why?
>
> Another question is, when I have successfully generate a psf file from the structure in which atoms are segmented into different segments, wouldn't they affect each other's secondary struc guessing even if the atoms are placed badly, or would they?
>
>
>
> In addition, here is the error messages from VMD console:
> ERROR) Unable to find Stride output file:
> C:\Users\CCHAN2~1\AppData\Local\Temp\3
> ERROR) Stride::read_stride_record: unable to read output file from Stride
> ERROR) Call to Stride program failed.
> And it gives me a pop-out window saying stride_WIN32.exe has stopped
> working.
>
> Thanks in advance,
> Kevin
>
> The pop-up window indicating the STRIDE stopped means that it crashed
> for some reason, and this is why VMD isn't getting output from it.
> I don't have a particular suggestion for you at this moment, but it might
> be useful for me to know more details about your structure such as how large
> it is, etc.
> Cheers,
> John Stone
> vmd_at_ks.uiuc.edu <mailto:vmd_at_ks.uiuc.edu?Subject=Re:%20%20Problem%20with%20assignation%20of%20secondary%20structure%20in%20VMD>
> On Fri, May 08, 2015 at 11:50:35AM +0800, Kevin C Chan wrote:
> > Thanks for the reply.
> >
> > It says:
> > ERROR) Unable to find Stride output file:
> > C:\Users\CCHAN2~1\AppData\Local\Temp\3
> >
> > ERROR) Stride::read_stride_record: unable to read output file from Stride
> > ERROR) Call to Stride program failed.
> >
> > And it gives me a pop-out window saying stride_WIN32.exe has stopped
> > working.
> >
> > Regards,
> > Kevin