Yes, when building in MIR maps you should always start building the CA atoms into a dummy poly-Ala molecule (see Q.075). Use either the Baton or the (older) Bone_pick_CA command to do this. Use the Slider commands to make guesses concerning the residue types, or do it "by eye". On the latter subject, Freddy Vellieux wrote:
"If you wish to do it 'visually', try to locate in your sequence some clearly recognizable patterns. For this purpose, very large side chains are quite useful. For example, a sequence TRP-ALA-GLY-TYR can be quite interesting: TRP is very big; this will be followed by 2 residues with ~ no side chain density, then a big chunk of density, but not as voluminous as the TRP. Long charged side chains can be flexible, so beware: the density may not be as long as you'd expect it to be.
Another help in finding out the direction of your chain is from alpha helical segments: the vector C-alpha C-beta is directed ~ towards the N-terminal end of the helix. By large, most side chains are directed ~ towards this end."
(The latter phenomenon gives rise to the typical "Christmas tree" shape of the density in helical segments.)