[ccp4bb]: modelling multiple conformation

Dear All,

My question concerns a special case of modelling multiple conformations. I am refining a protein structure with the latest version of Refmac. At some point in the structure I can see two conformations for two residues (His, Ala). The Ca coordinates from one conformation continue nicely into the Ca backbone of the structure. However the Ca coordinates of the other two residues don't, that means the carbonyl atom of the Ala does not join with the nitrogen of the next residue with normal single conformation.

So my problem is: I can easily model two conformations for the first residue with a double conformation, the His. However I then also have two conformations for the next Ala. One Ala (Ala A) is after the His (His A) in one conformation, the other Ala (Ala B) is after the His (His B) in the other double conformation. Ala B does not continue into the next residue in the protein, but basically forms an end, whereas Ala A continues normally into the next residue of the protein. The e-densities for the two Ala are in very differnent spacial positions (I have some autoprocessing going on in my protein). What Refmac now appears to be doing is to move Ala B after His A, in a position close to Ala A, in order to join the carbonyl of Ala B with the nitrogen of the next residue in the protein.

What do I have to do that Refmac accepts and refines the Ala B in the position I have modelled it in, without moving it back to join the rest of the protein?

(Additionally I also have a multiple conformation for Ala A (i.e. actually triple conformation), but this one joins the next residues normally, so this one should not be a problem as long as Ala B stays were I have put it.)

Thanks very much,

Florian