Dear Yuri,

There is a general rule that you need data to 1.2 Angstroms or better to
solve a structure by ab initio direct methods. To be more precise, more
than half the reflections between 1.1 and 1.2 A should have I>2sigma(I).
So you are probably just within this limit. However it appears that the
largest structures solved this way with no atom heavier than S were not
more than half the size of your protein. The programs most often used
for such crazy attempts are SnB, SIR and SHELXD. It would be a good idea 
to find a workstation with at least 32 CPUs and run the multi-CPU version
of SHELXD and be patient, it might take a few weeks.

If you just want to remove model bias (a serious problem with MR
solutions using 3A data, but no problem with 1.15A data) you could try
MRSAD phasing. There are many good programs for doing this, but a 
particularly simple way would be to use ANODE to find the anomalous
sites from your MR solution and the new beta-test SHELXE for the
density modification and tracing. The only other program you would
need for this is SHELXC. I am still developing these programs but am
happy to provide them on email request. No other programs, libraries
etc. are needed because these programs have zero dependencies.

Best wishes, George
On Fri, Sep 23, 2011 at 07:49:33PM +0100, Yuri Pompeu wrote:
> Hello everyone,
> I have a data set >99% completeness to 1.15A
> This is a 400 amino acid long protein and it has 7 Met (Sulfur peaks around 
> 20sigma)
> And a tightly bound phosphate (P peak around 22sigma)
> Could I try and solve this directly or is it crazy idea?
> If so what program should I try?
> 
> thanks 
> Yuri
> 
> 

-- 
Prof. George M. Sheldrick FRS
Dept. Structural Chemistry, 
University of Goettingen,
Tammannstr. 4,
D37077 Goettingen, Germany
Tel. +49-551-39-3021 or -3068
Fax. +49-551-39-22582

Reply via email to