>likely oxidized
Yes I was forgetting we don't know what the redox state is after the first few xray exposures.
On 03/21/2016 03:57 PM, Gauss, George wrote:
> Thanks Edward. The FE is likely oxidized (i.e. crystals were grown aerobically without reductant). The FE is pentacoordinate with no evidence of FoFc density in the second axial coordination position.
>
> ________________________________________
> From: Edward A. Berry <[log in to unmask]>
> Sent: Monday, March 21, 2016 12:49 PM
> To: Gauss, George; [log in to unmask]
> Subject: Re: [ccp4bb] Metal bond length discrepancy
>
> For Phenix I think the most direct way is to look in the .geo file produced at the beginning of the run. This lists all the restraints, together with starting values. They are listed with worst offenders first, so if you look at the .geo file from refining the refmac model in phenix, it should be near the top.
>
> Is the heme pentacoordinate or hex, if the latter what is the sixth ligand? Is it oxidized or reduced? I think that for for oxidized hemes with two axial His ligands, ~2.0 is normal. Of course you know the story about O2 binding hemoglobin and pulling the Fe into the plane of the heme- the bond lengths vary between oxy and deoxy, so not sure one length fits for all.
>
> On 03/21/2016 02:05 PM, Gauss, George wrote:
>> My model has a Heme ligand with a His NE2 atom occupying the axial coordination position of the Heme FE. REFMAC refines the FE-NE2 bond length to approximately 1.9 Angstroms. PHENIX refines the same FE-NE2 bond length to approximately 2.2 Angstroms.
>>
>> I assume the metal bond length discrepancy arises from a mismatch between the library files used by REFMAC and PHENIX. How do I verify which library the programs are using? I am running the current versions of both refinement programs. I run REFMAC from CCP4i version 6.5.019; and use the PHENIX GUI version dev-2341.
>>
>> George Gauss
>>
>
|