[gmx-developers] Possible Cryo-em Contribution

Berk Hess hess at cbr.su.se
Tue Feb 16 16:28:33 CET 2010


Such contributions are very welcome.
This sounds like a tool that is very useful to a relatively small amount
of people.

The problem is that Gromacs is currently not easily extensible.
Adding a method, such as the one you describe, requires changes in many
places in the code as well as changes to quite rigid file formats.
This also means that any functionality has to be maintained actively.
We would prefer the core Gromacs functionality to be relatively small
and efficient
and allow extensive functionality through modules.
We are planning (already for a long time, but now something is finally
to made the Gromacs code and file formats more modular.

I the mean time you could upload a package with additional files
and diff's to present files (or simply a complete source tree)
to the Gromacs website.


Paul C. Whitford wrote:
> Gromacs Developers,
>      We have implemented a flexible fitting algorithm to model atomic
> structures to cryo-em reconstructions (described in Orzechowski and
> Tama, Biophysical Journal 95(12) pp. 5692 - 5705.  This is a
> different method than the one in NAMD) into Gromacs 4.0.5.  I was
> told that it is possible to contribute the code, if approved.   I was
> also informed that in order to contribute it, I should implement the
> code into the GIT version and ensure that all standards are followed
> (as described on the website, in addition to including all settings
> and map into the tpr file).  This will not be a problem.
> We have tested the software and we can rapidly fit systems with large
> conformational differences (~50 A rearrangements).  We have tested the
> code with the Amber and opls forcefields and with a simplified, all-atom,
> forcefield  (http://smog.ucsd.edu, the paper describing the webtool is
> currently under review).
> Please let me know if contribution of this method to the standard Gromacs
> release is of interest.   If it is, I will implement the code into a Git
> branch (provided one is created) and make sure it conforms to all other
> Gromacs standards.
> Thanks
> Paul C. Whitford
> Sanbonmatsu Team
> Theoretical Biology and Biophysics Group
> Los Alamos National Laboratory

More information about the gromacs.org_gmx-developers mailing list