[ccp4bb] r-factor does not reduce

2009-09-10 Thread Sylvia Fanucchi
Hi everyone I am new to solving crystal structures so apologies for the simple question I am solving a structure using molecular replacement. I have refined my structure with refmac however the R-factor and R-free do not seem to go down with successive rounds of refinement. The initial

Re: [ccp4bb] r-factor does not reduce

2009-09-10 Thread Sean Seaver
Hello Sylvia, Two issues I would consider: 1) recheck the space group 2) make sure that I had selected a good model for my MR search All the Best, Sean

Re: [ccp4bb] r-factor does not reduce

2009-09-10 Thread Roger Rowlett
It would not necessarily be unusual for an initial MR solution to have an R-factor of 45-48%, depending on the MR program used and quality of the search model. However, after an initial restrained refinement in Refmac, a good solution should drop pretty smartly from there. Some issues to

Re: [ccp4bb] r-factor does not reduce

2009-09-10 Thread Ezra Peisach
On 09/10/2009 12:17 PM, Sean Seaver wrote: Hello Sylvia, Two issues I would consider: 1) recheck the space group 2) make sure that I had selected a good model for my MR search All the Best, Sean Hi Sylvia, If you have just used rigid body refinement - the Rfactor might be high until

Re: [ccp4bb] r-factor does not reduce

2009-09-10 Thread Pavel Afonine
Hi Sylvia, although modern MR programs perform rigid body refinement, it may be worth of giving a try a novel rigid body refinement protocol implemented in phenix.refine (MZ rigid-body refinement protocol), that has very high convergence radius: J. Appl. Cryst. (2009). 42, 607-615. "Automatic

Re: [ccp4bb] r-factor does not reduce

2009-09-10 Thread Anastassis Perrakis
On 10 Sep 2009, at 19:29, Pavel Afonine wrote: Hi Sylvia, although modern MR programs perform rigid body refinement, it may be worth of giving a try a novel rigid body refinement protocol implemented in phenix.refine (MZ rigid-body refinement protocol), that has very high convergence