AMBER Archive (2009)

Subject: Re: [AMBER] RDC analysis: "No restraint defined"

From: case (case_at_biomaps.rutgers.edu)
Date: Wed Sep 16 2009 - 11:47:41 CDT


On Wed, Sep 16, 2009, Michael Durney wrote:

> I am trying to use AMBER 10 to fit RDC data to fixed coordinates but
> am having a problem with a "No restraint defined" message. Any help
> would really be appreciated!

The "no restraint defined" message is bogus -- it just means that you have
no "regular" (e.g. distance) NMR restraints defined...sander doesn't consider
RDC restaints when printing out that message. So, you can ignore that
(although we need to fix that up in the next release).

> Even though the RDC is not defined my output file does somehow list my
> RDC data in the final results but without fitting it:

This is the result of another buglet in the code: if the minimization stops
before reaching the maximum number of cycles, the summary prints out results
corresponding to the *first* step, not the last one. Since you had an initial
alignment tensor of all zeroes (which is fine), the program is printing
under "current" what was really there at the first step.

The workaround is to rerun, setting maxcyc to 55 (say, for your example).
Then the output should correspond to the final step, not the first step.
Of course, we need to fix this bug as well -- I have a vague recollection
that something tricky was going on that prevented me from doing that earlier,
but I'll look at it again. But see if the workaround helps.

...thanks for the report...good luck...dac

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber