[phenixbb] command line molprobity tools

Nathaniel Echols nechols at lbl.gov
Wed May 1 12:51:45 PDT 2013


On Wed, May 1, 2013 at 12:39 PM, Engin Özkan <eozkan at stanford.edu> wrote:
> I saw mmtbx.validation_summary, but it says it extracts header info.

Only R-factors and bond and angle RMSDs (and the latter only for files
from phenix.refine).  The actual MolProbity stuff is calculated fresh.
 It's not a very sophisticated tool, I just use it to benchmark
massive sets of test refinements (where I always run Phenix at the end
with a null strategy, to control for differences in bulk solvent and
hydrogen handling).

> That is
> not running molprobity for a file that came from a different refinement
> program.

Why would you ever want to run one of those? :)

> Comprehensive, but not the whole kitchen sink, would be what you see in the
> GUI reported, but in simple text form. The percentages/numbers of outliers,
> followed by the identity of those outliers. It would be great if when I hit
> the button for "View Log File" (or more accurately, a "View Result File"),
> what would appear is what I see in the GUI, which is concise enough but also
> detailed at the perfect level.

We have discussed making a unified program ("phenix.molprobity") that
would do something like this.  There aren't any technical obstacles;
the only reason it hasn't happened is that this really needs to happen
as a part of a more comprehensive refactoring of both the validation
tools in the Phenix GUI, and of the MolProbity server itself.
Otherwise we will inevitably end up with another jumbled mess with
lots of code duplication.

However, in the absence of that, I will add a --verbose flag to
mmtbx.validation_summary which will dump the output of the individual
programs sequentially.  It won't be especially pretty, but it's only a
handful of lines of code.

-Nat


More information about the phenixbb mailing list