Dear HK,

If you believe that R120 only adopts one conformation then you should not refine the occupancies of the side chain atoms. It makes no physical sense to have occupancies less than one for these atoms, but occupancy equal to one for the rest of the protein (unless you believe the side chain to be cleaved, or otherwise chemically modified). However, occupancy refinement of the ligand itself can be justified by partial binding.

Best regards,
Rob


> On 5 Nov 2019, at 15:55, [log in to unmask] wrote:
> 
> Dear Rob,
> 
> I would like to model the alternative position for the side chain of R120 but I don't really know whether the alternative conformation exist as shown in the attached figure - density without the ligand and R120 overlaid with the refined structures of modeled ligand and R120. The ligand was modeled in two different conformations. From the density, it seems to me that the density is connected or overlapped.
> 
> It should not be a post-translation modification as it is well-known that there is no post-translation modification for this protein. Furthermore, the crystal was obtained by co-crystallization of protein with the ligand itself. The density seems to be the expected ligand.
> 
> Thanks for the advice.
> 
> Best regards
> HK
> 
> 
> Quoting Robert Nicholls <[log in to unmask]>:
> 
>> Dear HK,
>> 
>> No that's not quite correct - 'occupancy group alts complete' means that both R120 and the ligand are constrained so that their occupancies sum to unity. In contrast, 'occupancy group alts incomplete' means that the occupancies of R120 and the ligand will not be constrained to sum to unity (but the sum of their occupancies must be less than one). In both cases, R120 and the ligand will "see" each other in a certain sense. But, because they are assigned to different groups, it is assumed that they are present in different parts of the crystal. This means that they can overlap.
>> 
>> Assuming that the ligand is in the correct conformation, I suspect the source of your problem is that you are modelling the side chain of R120 as only one conformation. And I would also include the other atoms in the side chain - CB and CG.
>> 
>> If you are modelling the sidechain of R120 with partial occupancies, then you should model those side chain atoms in two alternative positions (i.e. representing the portions of the crystal that do/don't have the ligand bound). This will help to ensure that your model makes physical sense. So the ligand plus the alt of R120 in the portion of the crystal that contains the ligand would be assigned to one occupancy group, and the alt of R120 in the portion of the crystal that does not contain the ligand would be assigned to the second group. In this case it would be appropriate to specify 'occupancy group alts complete', because the occupancies for the two alternative conformers of R120 should sum to unity. Although no doubt the estimation of the occupancies would be dominated by the ligand in this case.
>> 
>> Be sure to check your B-factors after occupancy refinement to make sure the whole picture makes sense. Assuming your current model is essentially correct, from visual inspection it looks like R120 will have low occupancy and low B-factors when the ligand is not present (or at least B-factors consistent with the environment), but will have high occupancy and high B-factors when the ligand is present.
>> 
>> On another note, have you considered whether that part of the ligand could be modelled in a slightly different conformation, or whether there could be a post-translation modification?
>> 
>> I hope that helps,
>> Rob
>> 
>> 
>> Dr Rob Nicholls
>> Senior Investigator Scientist
>> MRC Laboratory of Molecular Biology
>> Francis Crick Avenue
>> Cambridge Biomedical Campus
>> Cambridge CB2 0QH
>> 
>> 
>> 
>>> On 5 Nov 2019, at 13:36, HK <[log in to unmask]> wrote:
>>> 
>>> Dear all,
>>> 
>>> I have problem with occupancy refinement by Refmac5 for an overlapping electron density of part of residue (an arginine) and part of ligand (tetracyclic compound) (attached figures in Dropbox with a link as shown below).
>>> 
>>> https://www.dropbox.com/sh/ppmfp5dnpy1b9e9/AAAV79bOzPHQUrVYp9loMwyha?dl=0
>>> 
>>> I refined part of the side chain of residue 120 and ligand (chain J residue 1105) with Refmac keyword as shown below. Unfortunately, the side chain of arginine moves away from the density but the ligand stays in the density. As far as I understood, occupancy refinement with keyword 'occupancy group alts complete' means both R120 and the ligand do not meet each other.  Did I miss something from the occupancy refinement keyword?
>>> 
>>> occupancy group id 1 chain A residue 120 atom NE
>>> occupancy group id 1 chain A residue 120 atom CZ
>>> occupancy group id 1 chain A residue 120 atom NH2
>>> occupancy group id 1 chain A residue 120 atom NH1
>>> occupancy group id 1 chain A residue 120 atom CD
>>> occupancy group id 2 chain J residue 1105
>>> occupancy group alts complete 1 2
>>> occupancy refine
>>> 
>>> Thank you for the advice.
>>> 
>>> Best regards
>>> HK
>>> 
>>> ########################################################################
>>> 
>>> To unsubscribe from the CCP4BB list, click the following link:
>>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
>> 
>> 
>> ########################################################################
>> 
>> To unsubscribe from the CCP4BB list, click the following link:
>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
> 
> 
> <unmodeled_density.png><unmodelled_density2.png>

########################################################################

To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1