Hello
I have two questions relating to the same protein build. I am struggling with
an interesting case of a protein structure for which we have chosen to build it
with microheterogeneity. In summary, we have an all alpha helical highly
repetitive solenoidal RNA binding protein. In our structure half the protein is
represented in the ASU, and the other half is represented by crystallographic
symmetry. It features an infinite superhelix running the length of the crystal,
similar to that seen previously for work from our group (PDBid: 4OZS). The
protein is an alpha solenoidal structure, made up of helix/turn/helix motif,
that stack together to form a superhelix. In our crystal, these form infinite
superhelices that have no obvious beginning or end owing to a shift in the
“phase” of the helix as you move from one strand to another. Please refer to
this paper for a more in depth explanation (Acta Crystallogr. D Biol.
Crystallogr. 71 196-208 (2015)).
Ultimately we choose to build the model (2 Å data set) such that two residues
in each 35 residue helix/turn/helix motif (there are 18 motifs in total, 9 in
each ASU) have different amino acids at these two positions in the structure. I
represent these just as I would for alternate conformations, using A and B
switches to stipulate the different conformations. However, in our case, they
are in fact different residues. Now I can get COOT to visualise these well
enough by putting the following line in my .coot.py file;
allow_duplicate_sequence_numbers()
However, at this stage, REFMAC will fail and give me the following error;
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 5
different residues have the same number
ERROR: in chain A residue: 35
different residues have the same number
ERROR: in chain A residue: 35
different residues have the same number
And so on….
Question 1: How can I get REFMAC to accommodate microheterogeniety?
Question 2: As an additional question, how do I create a LINK of the N-terminus
of my model to the C-terminus of the neighbouring symmetry object as part of
these infinite superhelices? Such that it would be recognised by REFMAC during
refinement as a ‘continuous’ structure.
I appreciate any advice that is offered.
Yours,
Jason Schmidberger
