On 12/01/2010 12:39 AM, Anna Marabotti wrote: > sorry for the missed attachments. However, in the mean time I found the > solution by myself: in the PDB file there was a "TER" after the breaking of > the chain, therefore probably the program interpretates (correctly!) this as > the signal for a new chain (labelled with the same chain ID as the previous > one) instead as of a chain breaking. After removing the TER line, the script > worked properly.
Modeller has no problem with proteins containing multiple chains with the same ID - the warning is there simply because it will be hard to refer to the residues or atoms uniquely. For example, if you have two chains both with ID 'A', then mdl.residues['10:A'] will give you residue 10 in the *first* chain labeled 'A'. There is no way to get residue 10 in the *second* chain labeled A using this syntax; to distinguish them you'd have to refer to the chains numerically instead, i.e. mdl.chains.residues['10'] and mdl.chains.residues['10']
Since the mutate model script in the wiki uses the first syntax, it won't be able to find residues in the second chain with the same ID.
Ben Webb, Modeller Caretaker