                   STEM-GROUP POSITIONING

    Activating the 'Stem-group  Position'  item  of  the  3D
'Edit'  pulldown  menu  allows one to interactively define a
group of stems which may then be translated and rotated as a
rigid  body.   Rotation  is about any of the three axes of a
Cartesian frame located at the group's center of  mass,  and
translation is along these same axes.

   The purpose of this type of  editing  originated  from  a
pseudoknot-related  problem.  By default, our 2D model of an
RNA molecule assumes that a pseudoknot has its participating
stems coaxially stacked.  As became evident for a particular
case,  this  may  not  be  appropriate  for   all   proposed
pseudoknots.   Hence the need to undo the stacking. But this
unstacking is not easy  to  accomplish  by  editing  the  2D
model, -as when unstacking or stacking non-pseudoknot stems.
Moreover, it appears to introduce stereochemically incorrect
distortions  that  argue  for  conformations intermediate to
those of the fully unstacked case. These  intermediates  are
better achieved by doing editing at the 3D level.  Hence the
requirement  of   moving  one  of  the  two  stems  of   the
pseudoknot  relative  to  the  other  as  a rigid body.  And
because in this particular case the loop of one of the stems
also  contained  two other stems, it proved necessay to also
consider moving these as a group of rigid bodies in  concert
with  and independent of the unstacking. These motions could
not be achieved in terms of segment repositioning  (see  the
help  topic  SEGMENT  POSITIONING),  and  so the new editing
feature.

   A member of a stem_group can only be a  hairpin  (a  non-
branching  stem  and  its  hairpin  loop)  or  a  stem  of a
pseudoknot. This restriction is necessary for the editing to
be applicable to pseudoknots.

   An example of usage is with  regard  to  the  sample  BPL
"hcvirus".   It  has  a total of four stems.  The yellow and
red colored stems constitute a proposed pseudoknot, and  the
two  that  are  cyan colored belong to the branching loop of
the red colored  stem.   Invoking  the  STEM_GROUP  POSITION
feature  and  picking  the  yellow  stem  as  the  one to be
repositioned (it becomes green upon picking), it is  readily
evident  that  unstacking  can  be  accmplished by combining
translation along the white and blue axes.   Some  of  these
translations  will  suggest  that the two cyan colored stems
should be moved in concert with the  yellow  stem.   Picking
these  will  add  them  to  the "stem_group".  It is also of
interest to observe that  the  cyan  colored  stems  can  be
stacked  to  each  other.   This  can only be done at the 2D
level and must precede  generation  of  the  3D  model.   So
invoke the 2D editing option "Stem Stacking" and stack these
two stems.  The 3D model will automatically  be  regenerated
and one can then continue with the 3D editing,- although the
stem_group  must  first  be  redefined.   During  all  these
motions   it   will  be  noticed  that  the  single  strands
connecting the stem_group to the rest of the  molecule  tend
to  be  adjusting their conformation to the new positions of
the  stem_group.   This  is  because  at  the  end  of  each
translation  or rotation step there is automatically invoked
a refinement of these strands at the same level that is used
at  the  end  of the 2D to 3D conversion process (Help topic
PRE-REFINEMENT).

   Needless  to  say,  the  "stem-group  position"   editing
feature   is   only   for   achieving   alternate   starting
conformations.  As in the unedited case, these conformations
should  be  regarded as first-order approximations to stable
ones, and hence the need for subsequent refinement.

                          THE END




