structure refinement

Gerard 'CD' Kleywegt GERARD at XRAY.BMC.UU.SE
Wed May 8 08:40:41 EST 1996

technically it's very simple, but i don't think you really want to
do this !  if rfree stays high, it indicates there is a problem,
either with the model or the data (or both) !  it's much more
useful to try and pin down this problem rather than to try and
fudge rfree (imho)


if you insist:

before the slow cool do:

   { initialise "best Rfree" variable }
   evaluate ($bestr = 1.0)
   { store current coordinates }
   coord copy end

inside the slow cool loop do:

     { store structure with lowest Rfree }
     xrefin update print r end
     if ($test_r LE $bestr) then
       coord copy end
       evaluate ($bestr = $test_r)
     end if

after the slow cool do:

   { restore structure with lowest Rfree }
   coord swap end
   xrefin update print r end

you'll probably want to do some room-temp MD (since
you're essentially doing a fast cool, going from perhaps
3500 K down to 300 K in one go)

More information about the X-plor mailing list