There is a great confusion over this nomenclature.
R3 in the rhombehedral setting has a=b=c, and alpha = beta = gamma.
H3 - the hexagonal setting has a=b, and alpha=beta=90 and gamma = 120.
Most CCP4 programs I believe ignore your named spacegroup and look at
the cell dimensions to make an intelligent choice!
But you need to do the editting yourself and give a name consistent with
your choice of cell
OK?
Eleanor
Shankar Prasad Kanaujia wrote:
> Dear all,
> I am having a problem in validating my refined model with ADIT server from
> RCSB. The problem is actually due to the space group.
> Actually the data was processed in R3 space group. So, there are two
> conventions for this space group (Hexagonal and Rhombohedral). I refined
> the model with R3 using CNS because H3 is not acceptable by CNS. But,
> actually the space group is H3 which I confirmed while checking for the
> twinning using UCLA server and also when i tried to generate the symmetry
> related molecules in COOT, it was not able to generate symmetry related
> molecules properly (all bonds were broken). But when I replaced R3 with
> H3, COOT was able to generate the symmetry related molecules properly
> with no short contacts. Now when I wanted to validate the
> final refined model using ADIT server (RCSB) with R3 space group, it gives
> some close contacts with symmetry related molecule. When I checked in COOT
> using R3, there was short contact due to symmetry related molecule
> (remember the symmetry related was not properly generated by COOT in R3).
> Then I checked the map in FRODO and actually there was no short contact.
> So, I realized that actually the space group is H3 and not R3. And COOT
> also showed no short contact with H3 space group group (remember, in H3,
> COOT is able to generate symm. related molecule properly).
>
> The problem is that when I tried to validate the model with ADIT server
> with H3 space group, it is not able to do anything. Even after 24 hrs of
> ADIT run, there was no oupput (not even error in loading the files).
>
> When I checked the PDB statistics, actually there are 479 structures
> submitted in RCSB with H3 space group. So, I am very much sure that some
> (if not all) must have validated their refined model with ADIT in H3 space
> group. Did anybody face this kind of problem.
>
> For your kind information, my CRYST card is like this:
> CRYST1 125.608 125.608 277.323 90.00 90.00 120.00 H 3 9
>
> Any suggestions will be very helpful.
> Thanking you.
>
> ---
> Yours Sincerely,
> Shankar Prasad Kanaujia
> Research Student
> C/O - Dr. K. Sekar
> Bioinformatics Center, Department of SERC
> IISc, Bangalore - 12, INDIA.
> Office Phone: 2469, 3059
> Mobile: 9845631581
>
>
>
|