On Thu, 2011-03-03 at 14:13 +0800, Ting-Wei Jiang wrote:
> I'm trying to calculate R-value (and free R) specifically which is
> between data and the modified structure(refined by myself without help
> from any program).
At long last, someone escaped the tyranny and oppression of the
refinement programs and deciphered the message hidden within the data by
using the mystical power known to the ancients as "X-ray vision". Paging
Dan Brown.
On a serious note, if you have the Fo and Fc, the calculation is rather
trivial (if that is the case, let me know and I'll send you a little
piece of code that does that).
However, I suspect that all you have is the model (xyzb for every atom)
and Fo. In which case there are many ways to do the calulation.
- sfcheck
- refmac with 0 rounds of refinement (already mentioned by Eleanor)
- CNS script model_stats.inp
- clipper libs
- phenix.fmodel will probably output it too or maybe there is some
phenix.get_me_my_damn_rfactor or such
The problem you'll have is that values will be different with different
programs. Given that your model is exactly the same and presumably the
atomic scattering factors used by these programs are the same, the few
likely reasons are various scaling operations applied. IMHO, the main
source of difference should be the way one treats the bulk solvent
correction.
HTH, and yes, the first paragraph is meant to be a joke.
Ed.
--
"I'd jump in myself, if I weren't so good at whistling."
Julian, King of Lemurs
|