Hi,
The R-factor you mention, is it an R-factor before any refinement of the
model ? Like an R-factor at the very beginning of the entire modeling
procedure, right after molecular replacement ?
If this is so: you should always compare such initial R-factors to the
R-factors for the atoms placed randomly in the asymmetric unit (in the
crystal thus, see X-ray Structure Determination by Stout and Jensen, pp
246 in the edition I have here - 1968, 1st edition?). R for centric
reflections = 0.83, R for acentric reflections = 0.59.
In space group F4132 there are many centric reflections. The "random"
R-factor may be around 0.65 or higher... (I didn't count), and if the
value you quote (0.5) is indeed a "starting R-factor" it is well below
that for a random distribution of atoms in the asymmetric unit. Matthews
coefficient only gives a probability of the number of molecules in the
asymmetric unit, but it does not provide you with that figure. You may
well have plenty of solvent in your asymmetric unit.
The most important thing are the electron density maps: do these provide
indications of where the search model has to be modified in order to fit
the target molecule ?
Also, if you have a component missing from your molecular replacement
calculations (as in the case you locate only one molecule instead of
two), very often there is some additional "proteinacious" density
appearing in the crystal contact regions between your correctly placed
molecule and the "ghost" molecule (which you haven't positioned yet).
So if R=0.5 is the starting R-factor just after molecular replacement, I
think it is an appropriate R-factor for molecular replacement calculations.
HTH,
Fred.
Afshan Begum wrote:
> Dear all,
>
> Could any one help me regarding my serious problem actually i have
> collected data at 3.0 and cut off 3.1 where the data statics showed
> the good values for the further processing.
> According to the methew coefficient there would be two molecule in
> the asymmetric unit but after running the molrep its provide only one
> monomer instead of two, for this reason R values is very high 50%.
> Actually homologous model having P6322 space group where as my one is
> F4132. I had tried to run phaser as well phenix but both were failed
> to process further. i really do not know how can i get the second
> chain in my structure. Please if you have some ideas i will appreciate
> and would be many thankful to you.
>
> Hope to hearing you soon
>
> Best Regards
>
> AFSHAN
>
> **
>
|